Information processing apparatus, information processing method, and program

Abstract

To achieve a technique to suitably derive a prediction result in an optimization method referring to a plurality of models.SOLUTION: An information processing apparatus comprises: a first acquisition unit that acquires data for inference and information on the class of the data for inference; a second acquisition unit that refers to the information on the class, specifies similar data that is data similar to the data for inference, and acquires a ratio parameter associated with the similar data; a prediction unit that applies a regression coefficient for each of a plurality of target models to the data for inference, and applies the ratio parameter associated with the similar data as a ratio parameter for defining the ratio between the plurality of target models, and thereby deriving a prediction result; and output means that outputs the prediction result derived by the prediction unit.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to an information processing device, an information processing method, and a program. [Background technology]

[0002] Optimization or control methods using multiple models are known. For example, Patent Document 1 discloses a federated learning system that learns a local model for each facility and trains the local models so as to reduce the difference between the models. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2023-182380 Summary of the Invention [Problem to be solved by the invention]

[0004] On the other hand, in addition to associative learning, a linear parameter-varying (LPV) model, which is expressed as a weighted sum of multiple models, is known as an optimization or control method that refers to the results of multiple models. In such a method that refers to multiple models, it is desirable to derive prediction results appropriately.

[0005] The present disclosure has been made in consideration of the above-mentioned problems, and an exemplary purpose thereof is to provide a technology that can suitably derive prediction results in an optimization method that references multiple models. [Means for solving the problem]

[0006] An information processing device according to an exemplary aspect of the present disclosure includes a first acquisition means for acquiring inference data and information relating to the type of the inference data, a second acquisition means for referring to the information on the type to identify similar data that is similar to the inference data and acquire ratio parameters associated with the similar data, a prediction means for applying regression coefficients for each of a plurality of target models to the inference data and applying the ratio parameters associated with the similar data as ratio parameters that define the ratios of the plurality of target models to derive a prediction result, and an output means for outputting the prediction result by the prediction means.

[0007] An information processing device according to an exemplary aspect of the present disclosure includes an acquisition means for acquiring target data, the number of models of a plurality of target models, and information relating to the type of the target data; a regression coefficient calculation means for calculating a regression coefficient for each of the plurality of target models by referring to the target data and a ratio parameter that defines a ratio of the plurality of target models; a ratio parameter calculation means for calculating the ratio parameter by referring to the target data and the regression coefficient; and a storage means for storing the regression coefficient calculated by the regression coefficient calculation means and the ratio parameter calculated by the ratio parameter calculation means in association with information relating to the type.

[0008] An information processing method according to an exemplary aspect of the present disclosure includes acquiring inference data and information relating to the type of the inference data, identifying similar data that is similar to the inference data by referring to the information on the type and acquiring ratio parameters associated with the similar data, applying regression coefficients for each of a plurality of target models to the inference data, and applying the ratio parameters associated with the similar data as ratio parameters that define the ratios of the plurality of target models to derive a prediction result, and outputting the prediction result.

[0009] An information processing method according to an exemplary aspect of the present disclosure includes acquiring target data, the number of models of a plurality of target models, and information relating to the type of the target data; calculating a regression coefficient for each of the plurality of target models by referring to the target data and a ratio parameter that defines a ratio of the plurality of target models; calculating the ratio parameter by referring to the target data and the regression coefficient; and storing the regression coefficient and the ratio parameter in association with information relating to the type.

[0010] The information processing device according to each aspect of the present invention may be realized by a computer, in which case a program that causes the computer to operate as each part (software element) of the information processing device to realize the information processing device also falls within the scope of the present invention. [Effects of the Invention]

[0011] According to an exemplary aspect of the present disclosure, an exemplary effect is achieved in that a prediction result can be suitably derived in an optimization method that references multiple models. [Brief explanation of the drawings]

[0012] [Figure 1] 1 is a block diagram illustrating a configuration of an information processing device according to the present disclosure. [Figure 2] FIG. 1 is a flow diagram showing the flow of an information processing method according to the present disclosure. [Figure 3] 1 is a block diagram illustrating a configuration of an information processing device according to the present disclosure. [Figure 4] FIG. 1 is a flow diagram showing the flow of an information processing method according to the present disclosure. [Figure 5] 1 is a block diagram illustrating a configuration of an information processing device according to the present disclosure. [Figure 6] FIG. 10 is a diagram illustrating an example of a processing flow in an information processing device according to the present disclosure. [Figure 7] 10 shows an example of a graph displayed by an output unit via an input / output unit according to the present disclosure. [Figure 8]10 shows an example of information that an output unit according to the present disclosure displays via an input / output unit. [Figure 9] 10 shows an example of a graph displayed by an output unit via an input / output unit according to the present disclosure. [Figure 10] 1 is a block diagram illustrating a configuration of an information processing device according to the present disclosure. [Figure 11] FIG. 10 is a diagram schematically illustrating the outputs of each endpoint model in the LPV model according to the present disclosure and the interior division ratio parameters by which each output is multiplied. [Figure 12] FIG. 10 is a diagram illustrating an example of a processing flow in an information processing device according to the present disclosure. [Figure 13] 10 shows an example of a graph displayed by an output unit via an input / output unit according to the present disclosure. [Figure 14] 1 is a block diagram illustrating a configuration of an information processing device according to the present disclosure. [Figure 15] FIG. 10 is a diagram illustrating an example of a processing flow in an information processing device according to the present disclosure. [Figure 16] FIG. 1 is a block diagram illustrating a configuration of a computer that functions as an information processing device according to the present disclosure. DETAILED DESCRIPTION OF THE INVENTION

[0013] The following are examples of embodiments of the present invention. However, the present invention is not limited to the exemplary embodiments shown below, and various modifications are possible within the scope of the claims. For example, embodiments obtained by appropriately combining the technical means employed in the exemplary embodiments shown below may also be included in the scope of the present invention. Furthermore, embodiments obtained by appropriately omitting some of the technical means employed in the exemplary embodiments shown below may also be included in the scope of the present invention. Furthermore, the effects mentioned in the exemplary embodiments shown below are examples of effects expected in the exemplary embodiments, and do not define the scope of the present invention. In other words, embodiments that do not exhibit the effects mentioned in the exemplary embodiments shown below may also be included in the scope of the present invention.

[0014] First Exemplary Embodiment A first exemplary embodiment, which is one example of an embodiment of the present invention, will be described in detail with reference to the drawings. This exemplary embodiment is the basic form of each exemplary embodiment described later. Note that the scope of application of each technical means employed in this exemplary embodiment is not limited to this exemplary embodiment. That is, each technical means employed in this exemplary embodiment can also be employed in other exemplary embodiments included in the present disclosure to the extent that no particular technical obstacles arise. Furthermore, each technical means shown in the drawings referred to in describing this exemplary embodiment can also be employed in other exemplary embodiments included in the present disclosure to the extent that no particular technical obstacles arise.

[0015] (Configuration of information processing device 1) The configuration of the information processing device 1 will be described with reference to Fig. 1. Fig. 1 is a block diagram showing the configuration of the information processing device 1. As shown in Fig. 1, the information processing device 1 includes a first acquisition unit 21, a second acquisition unit 22, a prediction unit 23, and an output unit 24. In this exemplary embodiment, the first acquisition unit 21, the second acquisition unit 22, the prediction unit 23, and the output unit 24 respectively realize a first acquisition means, a second acquisition means, a prediction means, and an output means.

[0016] (First acquisition unit 21) The first acquisition unit 21 acquires data for inference (target data) and information on the type of the data for inference. The first acquisition unit 21 supplies the acquired data for inference (target data) and information on the type to the second acquisition unit 22 and the prediction unit 23.

[0017] (Second acquisition unit 22) The second acquisition unit 22 refers to the information on the type acquired by the first acquisition unit 21, identifies similar data that is data similar to the inference data (target data), and acquires a ratio parameter associated with the similar data. Process 221: A process of selecting (identifying) similar data, which is data similar to the inference data (target data), from a plurality of candidate data by referring to the information about the type; and Process 222: Obtaining a ratio parameter associated with the similar data Here, each of the plurality of candidate data referred to in the process 221 is also associated with information on the type. The second acquisition unit 22 executes the process 221 by referring to the information on the type of the data for inference (target data) and the information on the type of each of the plurality of candidate data. As an example, the type of the data for inference (target data) is A, and the following candidate data is selected: Candidate data 1: Type a (similar to Type A) Candidate data 2: Type b (not similar to Type A) Candidate Data 3: Type c (not similar to Type A) is given, the second acquiring unit 22 identifies candidate data 1 having a type similar to the type of the data for inference (target data) as the similar data.

[0018] Furthermore, in the process 222, the "ratio parameter associated with similar data" refers to, as an example, a ratio parameter obtained by a learning process that references the similar data, but this does not limit this exemplary embodiment.

[0019] (Prediction Section 23) The prediction unit 23 derives a prediction result by referring to the inference data and the ratio parameter. Process 231: A process of applying a regression coefficient for each of a plurality of target models to the inference data. Process 232: A process of applying the ratio parameter associated with the similar data to the inference data as a ratio parameter that defines the ratio of the plurality of target models. Here, the regression coefficients for each of the plurality of target models refer to, for example, regression coefficients of the plurality of target models trained with reference to the similar data, but this does not limit the present exemplary embodiment.

[0020] As an example, in process 231, the prediction unit 23 Step 2311: For the feature x0 included in the inference data, a regression coefficient W for the first target model (a regression coefficient possessed by the first target model) is calculated. (1) By applying (multiplying) (1) =W (1) Calculate ·x0 (where "·" represents a product or dot product), Step 2312: For the feature x0 included in the inference data, a regression coefficient W for the second target model (a regression coefficient possessed by the second target model) is calculated. (2) The second predicted value y (2) =W (2) Calculate x0.

[0021] Then, in process 232, the prediction unit 23 applies the ratio parameter associated with the similar data to the results (the first predicted value, the second predicted value) of applying the regression coefficients of the respective target models (the first target model, the second target model) to the inference data as a ratio parameter that defines the ratio of the plurality of target models. As an example, in process 232, the prediction unit 23 Process 2321: Ratio parameter μ associated with the similar data (1) and μ (2) Among these, μ is a ratio parameter that defines the ratio of the first target model (in other words, a ratio parameter that defines the ratio of the first predicted values). (1) the predicted value y (1) By applying (multiplying) to μ (1) ·y (1) Calculate Process 2322: Ratio parameter μ associated with the similar data (1) and μ(2) Among these, μ is a ratio parameter that defines the ratio of the second target model (in other words, a ratio parameter that defines the ratio of the second predicted values). (2) the predicted value y (2) By applying (multiplying) to μ (2) ·y (2) Calculate Process 2323: Using the results of processes 2321 and 2322, the prediction result is calculated using μ (1) ·y (1) +μ (2) ·y (2) Derive it as However, the above processing example does not limit this exemplary embodiment.

[0022] (output unit 24) The output unit 24 outputs the prediction result derived by the prediction unit 23. For example, the output unit 24 visually presents the prediction result to the user via a display means (not shown). The output unit 24 may also store the prediction result in a storage means (not shown).

[0023] (Effects of information processing device 1) As described above, in the information processing device 1, Acquire inference data and information regarding the type of the inference data, Identifying similar data that is similar to the inference data by referring to the information about the type, and obtaining a ratio parameter associated with the similar data; The inference data includes: Applying the regression coefficients for each of a plurality of target models; The ratio parameter associated with the similar data is applied as a ratio parameter that defines a ratio of the plurality of target models. The prediction results are derived by The information processing device 1 has the above configuration. Therefore, the information processing device 1 can preferably derive a prediction result by utilizing a ratio parameter associated with similar data, which is data similar to the inference data. In other words, the information processing device 1 can preferably derive a prediction result in an optimization method that refers to multiple models.

[0024] (Flow of information processing method S1) The flow of the information processing method S1 will be described with reference to Fig. 2. Fig. 2 is a flow diagram showing the flow of the information processing method S1. As shown in Fig. 2, the information processing method S1 includes a first acquisition process S21, a second acquisition process S22, a prediction result derivation process S23, and a prediction result output process S24.

[0025] (First acquisition process S21) In the first acquisition process S21, the first acquisition unit 21 acquires data for inference (target data) and information on the type of the data for inference. The acquisition unit 21 supplies the acquired data for inference (target data) and information on the type to the second acquisition unit 22 and the prediction unit 23.

[0026] (Second acquisition process S22) In the first acquisition process 22, the second acquisition unit 22 refers to the information on the type acquired by the first acquisition unit 21, identifies similar data that is data similar to the inference data (target data), and acquires ratio parameters associated with the similar data. The specific process by the second acquisition unit 22 has been described above, so a description thereof will be omitted here.

[0027] (Prediction result derivation process S23) In the prediction result derivation process S23, the prediction unit 23 derives a prediction result by referring to the inference data and the ratio parameter. Process 23-1: A process of applying a regression coefficient for each of a plurality of target models to the inference data. Process 23-2: Applying the ratio parameter associated with the similar data to the inference data as a ratio parameter that defines the ratio of the plurality of target models. The prediction result is derived by executing the process described above. Here, the regression coefficients for each of the plurality of target models refer, as an example, to the regression coefficients of the plurality of target models trained with reference to the similar data, but this does not limit the present exemplary embodiment. The specific process performed by the prediction unit 23 has been described above, so a description thereof will be omitted here.

[0028] (output unit 24) In a prediction result output process S24, the output unit 24 outputs the prediction result derived by the prediction unit 23. The specific process by the output unit 24 has been described above, and therefore will not be described here.

[0029] (Effect of information processing method S1) As described above, in the information processing method S1, Acquire inference data and information regarding the type of the inference data, Identifying similar data that is similar to the inference data by referring to the information about the type, and obtaining a ratio parameter associated with the similar data; The inference data includes: Applying the regression coefficients for each of a plurality of target models; The ratio parameter associated with the similar data is applied as a ratio parameter that defines a ratio of the plurality of target models. The prediction results are derived by Therefore, the information processing method S1 has the same effect as the information processing device 1.

[0030] (Configuration of information processing device 2) The configuration of the information processing device 2 will be described with reference to Fig. 3. Fig. 3 is a block diagram showing the configuration of the information processing device 2. As shown in Fig. 3, the information processing device 2 includes an acquisition unit 11, a regression coefficient calculation unit 12, a ratio parameter calculation unit 14, and an output unit 15. In this exemplary embodiment, the acquisition unit 11, the regression coefficient calculation unit 12, the ratio parameter calculation unit 14, and the output unit 15 respectively realize an acquisition means, a regression coefficient calculation means, a ratio parameter calculation means, and a storage means.

[0031] (Acquisition part 11) The acquisition unit 11 acquires target data, the number of models of a plurality of target models, and information regarding the type of the target data. The acquisition unit 11 supplies the acquired target data to the regression coefficient calculation unit 12 and the ratio parameter calculation unit 14. The acquisition unit 11 also supplies the acquired number of models of the plurality of target models to the regression coefficient calculation unit 12. The acquisition unit 11 also supplies the acquired target data and information regarding the type of the target data to the output unit 15.

[0032] (Regression coefficient calculation part 12) The regression coefficient calculation unit 12 calculates a regression coefficient for each of the plurality of target models by referring to the target data and a ratio parameter that defines the ratio of the plurality of target models. The regression coefficient calculation unit 12 supplies the calculated regression coefficient to the ratio parameter calculation unit 14 and the output unit 15.

[0033] (Ratio parameter calculation unit 14) The ratio parameter calculation unit 14 calculates the ratio parameters by referring to the target data and the regression coefficients. Here, the ratio parameters are parameters that define the ratios of the multiple target models. The ratio parameter calculation unit 14 supplies the calculated ratio parameters to the output unit 15.

[0034] (Output section 15) The output unit 15 stores the regression coefficients calculated by the regression coefficient calculation unit 12 and the ratio parameters calculated by the ratio parameter calculation unit 14 in association with the information on the type. As an example, the output unit 15 supplies the regression coefficients and the ratio parameters to a storage unit (not shown) in association with the information on the type.

[0035] (Effects of information processing device 2) As described above, in the information processing device 2, Acquire information regarding the target data, the number of models of multiple target models, and the type of the target data; calculating a regression coefficient for each of the plurality of target models by referring to a ratio parameter that defines a ratio of the plurality of target models and the target data; Calculating the ratio parameter by referring to the target data and the regression coefficients; The regression coefficients and the ratio parameters are stored in association with information about the type. The information processing device 2 has the following configuration. Therefore, in an optimization method that refers to a plurality of models, information about the type is stored in association with the regression coefficients and the ratio parameters. Then, the information about the type can be suitably used in the prediction phase. Therefore, the information processing device 2 provides a technique for suitably deriving prediction results in an optimization method that refers to a plurality of models.

[0036] (Flow of information processing method S2) The flow of the information processing method S2 will be described with reference to Fig. 4. Fig. 2 is a flow diagram showing the flow of the information processing method S22. As shown in Fig. 4, the information processing method S2 includes an acquisition process S11, a regression coefficient calculation process S12, a ratio parameter calculation process S14, and a storage process S15.

[0037] (Acquisition process S11) In the acquisition process S11, the acquisition unit 11 acquires target data, the number of models of a plurality of target models, and information regarding the type of the target data. The acquisition unit 11 supplies the acquired target data to the regression coefficient calculation unit 12 and the ratio parameter calculation unit 14. The acquisition unit 11 also supplies the acquired number of models of the plurality of target models to the regression coefficient calculation unit 12. The acquisition unit 11 also supplies the acquired target data and information regarding the type of the target data to the output unit 15.

[0038] (Regression coefficient calculation process S12) In the regression coefficient calculation process S12, the regression coefficient calculation unit 12 calculates a regression coefficient for each of the plurality of target models by referring to the target data and a ratio parameter that defines the ratio of the plurality of target models. The regression coefficient calculation unit 12 supplies the calculated regression coefficient to the ratio parameter calculation unit 14 and the output unit 15.

[0039] (Ratio parameter calculation process S14) In the ratio parameter calculation process S14, the ratio parameter calculation unit 14 calculates the ratio parameters by referring to the target data and the regression coefficients. The ratio parameter calculation unit 14 supplies the calculated ratio parameters to the output unit 15.

[0040] (Amnestic S15) In the storage process S15, the output unit 15 stores the regression coefficients and the ratio parameters calculated by the ratio parameter calculation unit 14 in association with the information on the type.

[0041] (Effect of information processing method S2) As described above, in the information processing method S2, Acquire information regarding the target data, the number of models of multiple target models, and the type of the target data; calculating a regression coefficient for each of the plurality of target models by referring to a ratio parameter that defines a ratio of the plurality of target models and the target data; Calculating the ratio parameter by referring to the target data and the regression coefficients; The regression coefficients and the ratio parameters are stored in association with information about the type. Therefore, the information processing method S2 has the same effect as the information processing device 2.

[0042] Second Exemplary Embodiment A second exemplary embodiment, which is one example of an embodiment of the present invention, will be described in detail with reference to the drawings. Components having the same functions as those described in the above exemplary embodiment will be assigned the same reference numerals, and their description will be omitted as appropriate. The scope of application of each technical means employed in this exemplary embodiment is not limited to this exemplary embodiment. That is, each technical means employed in this exemplary embodiment can also be employed in other exemplary embodiments included in the present disclosure, to the extent that no particular technical obstacles arise. Furthermore, each technical means shown in each drawing referenced to explain this exemplary embodiment can also be employed in other exemplary embodiments included in the present disclosure, to the extent that no particular technical obstacles arise.

[0043] (Configuration of information processing device 1A) The configuration of the information processing device 1A will be described with reference to Fig. 5. Fig. 5 is a block diagram showing the configuration of the information processing device 1A. As shown in Fig. 5, the information processing device 1A includes a control unit 10A, a storage unit 15A, a communication unit 16A, and an input / output unit 17A.

[0044] (Storage section 15A) First, various data (information) stored in the storage unit 15A will be described. Data referenced by the control unit 10A or data derived by the control unit 10A is stored in the storage unit 15A. Examples of the storage unit 15A include, but are not limited to, a flash memory, a hard disk drive (HDD), a solid state drive (SSD), or a combination thereof.

[0045] Examples of data stored in memory unit 15A include, but are not limited to, inference data PD, type information CI, candidate data group CDG, learning results LR, and prediction results PR, as shown in FIG. 5.

[0046] (Inference data PD) The inference data PD is data to be inferred that is referenced in the inference process executed by the information processing device 1A. As an example, the inference data PD includes one or more feature quantities. The inference data PD may also be referred to as target data, and the feature quantities included in the inference data PD may also be referred to as internal state quantities or explanatory variables, but these names do not limit this exemplary embodiment.

[0047] (Type Information CI) The type information CI is information indicating the type of the inference data PD. Here, the information indicating the type may be, for example, information indicating the circumstances under which the inference data PD was acquired.

[0048] (Candidate Data Group CDG) The candidate data group CDG includes one or more candidate data CDs that are referenced in the inference process executed by the information processing device 1A. Here, each candidate data CD is, for example, Information about the type of candidate data CD One or more features Here, the information on the type of the candidate data CD may be, for example, information indicating the circumstances under which each candidate data CD was acquired. For example, when the one or more feature quantities include "day" as an explanatory variable, the information on the type may be, Candidate data CD1 (e.g., daily sales volume): Situation A (e.g., sunny weather) Candidate data CD2 (e.g., daily sales volume): Situation B (e.g., cloudy weather) Candidate data CD3 (e.g., daily sales volume): Situation C (e.g., the weather is rainy) Each candidate data CD may further include the following information: Ratio parameters obtained by the learning process referring to the candidate data CD Regression coefficients obtained by the learning process that references the candidate data CD The configuration may include the following.

[0049] In addition, information about the above types is available at Information indicating whether a certain candidate data is similar to other candidate data or data for inference Here, "a certain piece of data is similar to another piece of data" refers to, for example, a case where the type of the certain piece of data is similar to the type of the other piece of data. For example, if candidate data 1 is data on the number of sales acquired in situation A (e.g., sunny weather) as in the above example, candidate data or data for inference that is data on the number of sales in situation A' (e.g., sunny weather) similar to situation A is data similar to candidate data 1.

[0050] The storage unit 15A may be configured to further store correspondence information, which will be described below, as part of the information about the type, or together with the information about the type. The correspondence information is information about the correspondence between each candidate data CD and a plurality of target models. Examples of the information about the correspondence include: - Information indicating what kind of candidate data each of the multiple target models CD has been trained (will be trained) with is included. For example, the correspondence information regarding the target models a, b, and c and the candidate data CD1, CD2, and CD3 is as follows: Target model a: Candidate data CD1, CD2 Target model b: Candidate data CD3 Target model c: Candidate data CD1, CD2, CD3 Here, the first line in the above example indicates that the target model a has been trained (will be trained) using the candidate data CD1 and the candidate data CD2. The same applies to the other lines.

[0051] In other words, the correspondence information is Information indicating whether each of the multiple target models has used (will use) each of the multiple candidate data CD as training data As an example, the correspondence relationship information CI regarding the target models a, b, and c and the candidate data CD1, CD2, and CD3 may be expressed as follows: Target model a: Candidate data CD1 (○), CD2 (○), CD3 (×) Target model b: Candidate data CD1(×), CD2(×), CD3(○) Target model c: Candidate data CD1(○), CD2(○), CD3(○) Here, ○ indicates that it was used (will be used) in learning, and × indicates that it was not used (will not be used) in learning.

[0052] In addition, when the correspondence information includes the correspondence between each candidate data CD and the situation in which the candidate data CD was acquired as described above, the correspondence information may be expressed as including the relationship between each target model and the type related to the target model. For example, in the above example, the correspondence information is Target model a: Situation A (e.g., sunny weather), Situation B (e.g., cloudy weather) Target model b: Situation C (e.g., rainy weather) Target model c: Situations A, B, and C The correspondence information may include information indicating a correspondence such as the above. By including the correspondence information as described above, the information processing device 1A can identify what situation each target model is associated with, in other words, what situation each target model is preferably used in. Furthermore, the information processing device 1A can present to the user what situation each target model is preferably used in.

[0053] (Learning result LR) The learning result LR includes the results learned by the learning process with reference to each candidate data CD. For example, the learning result LR may include: Ratio parameters obtained by the learning process referring to each candidate data CD Regression coefficients obtained by the learning process referring to each candidate data CD Includes:

[0054] In one example, the regression coefficients and the ratio parameters are A regression coefficient calculation process that calculates a regression coefficient for each of a plurality of target models by referring to the ratio parameters and the candidate data CD; and A ratio parameter calculation process for calculating the ratio parameters by referring to the candidate data CD and the regression coefficients. are the regression coefficients and ratio parameters learned by a learning process including

[0055] Further, the learning process for the regression coefficients and the ratio parameters is A covariance calculation process that calculates the covariance parameters of the prior distribution of the latent variables and the covariance matrix of the posterior distribution of the latent variables by referring to similar data similar to the inference data PD, the ratio parameters, the regression coefficients, and information on the prior distribution of the latent variables. The ratio parameter calculation process may further include: Calculate the ratio parameter by further referring to the covariance matrix of the posterior distribution of the hidden variables. The processing may be as follows.

[0056] (Prediction results PR) The prediction result PR is a prediction result derived by the prediction unit 23, and is data to be output by the output unit 24. Details of the prediction result PR will be described later.

[0057] (Communication unit 16A) Communication unit 16A is an interface that transmits and receives data via a network. Examples of communication unit 16A include, but are not limited to, communication chips for various communication standards such as Ethernet (registered trademark), Wi-Fi (Wireless Fidelity (registered trademark), and wireless communication standards for mobile data communication networks, and USB-compliant connectors.

[0058] (Input / output section 17A) The input / output unit 17A is an interface that receives input of data and outputs data. Examples of the input / output unit 17A include, but are not limited to, a microphone, a camera, a gaze input device, a keyboard, a touchpad, a speaker, and a liquid crystal display.

[0059] (Control unit 10A) As shown in FIG. 5, the control unit 10A includes a first acquisition unit 21, a second acquisition unit 22, a prediction unit 23, and an output unit 24.

[0060] (First acquisition unit 21) The first acquisition unit 21 acquires inference data (target data) PD and information relating to the type of the inference data. The inference data and the information relating to the type have been described above, so a description thereof will be omitted here.

[0061] (Second acquisition unit 22) The second acquisition unit 22 refers to the information on the type acquired by the first acquisition unit 21, identifies similar data RD that is data similar to the inference data (target data), and acquires a ratio parameter associated with the similar data RD. Process 221: A process of selecting (identifying) similar data RD, which is data similar to the inference data (target data), from one or more candidate data CD included in the candidate data group CDG described above, by referring to the information related to the type; and Process 222: A process for obtaining a ratio parameter associated with the similar data RD Here, each of the plurality of candidate data CDs referenced in the process 221 is associated with information on the type as described above. The second acquisition unit 22 executes the process 221 by referring to information on the type of the data for inference (target data) PD and information on the type of each of the plurality of candidate data CDs. As an example, the type of the data for inference (target data) PD is A, and the storage unit 15A stores the candidate data CD. Candidate data CD1: Type a (similar to Type A) Candidate data CD2: Type b (not similar to Type A) Candidate data CD3: Type c (not similar to Type A) is stored, the second acquiring unit 22 identifies candidate data CD1 having a type similar to the type of the inference data (target data) PD as the similar data RD.

[0062] In the process 222, the "comparison parameter associated with the similar data RD" refers to, for example, a comparison parameter obtained by a learning process that references the similar data RD. Specific examples of the learning process are as described above. More specific processing by the second acquisition unit 22 will be described later.

[0063] (Prediction Section 23) The prediction unit 23 derives a prediction result PR by referring to the inference data PD and the ratio parameter. Process 231: A process of applying a regression coefficient for each of a plurality of target models to the inference data PD. Process 232: Applying the ratio parameter associated with the similar data RD to the inference data PD as a ratio parameter that defines the ratio of the plurality of target models. A prediction result is derived by executing the process of: Here, the regression coefficients for each of the plurality of target models refer to, for example, regression coefficients possessed by the plurality of target models trained with reference to the similar data, but this does not limit the present exemplary embodiment.

[0064] As an example, in process 231, the prediction unit 23 Step 2311: For the feature x0 included in the inference data PD, a regression coefficient W for the first target model (a regression coefficient possessed by the first target model) is calculated. (1) By applying (multiplying) (1) =W (1) Calculate ·x0 (where "·" represents a product or dot product), Step 2312: For the feature x0 included in the inference data PD, a regression coefficient W for the second target model (a regression coefficient possessed by the second target model) is calculated. (2) The second predicted value y (2) =W (2) Calculate x0.

[0065] Then, in process 232, the prediction unit 23 applies the ratio parameter associated with the similar data RD to the results (the first predicted value, the second predicted value) of applying the regression coefficients of the respective target models (the first target model, the second target model) to the inference data PD as a ratio parameter that defines the ratio of the plurality of target models. As an example, in process 232, the prediction unit 23 applies the ratio parameter associated with the similar data RD to the results (the first predicted value, the second predicted value) of applying the regression coefficients of the respective target models (the first target model, the second target model) to the inference data PD. Process 2321: Ratio parameter μ associated with the similar data RD (1) and μ (2) Among these, μ is a ratio parameter that defines the ratio of the first target model (in other words, a ratio parameter that defines the ratio of the first predicted values). (1) the predicted value y (1) By applying (multiplying) to μ (1) ·y (1) Calculate Process 2322: Ratio parameter μ associated with the similar data RD (1) and μ (2) Among these, μ is a ratio parameter that defines the ratio of the second target model (in other words, a ratio parameter that defines the ratio of the second predicted values). (2) the predicted value y (2) By applying (multiplying) to μ (2) ·y (2) Calculate Process 2323: Using the results of processes 2321 and 2322, the prediction result PR is calculated as μ (1) ·y (1) +μ (2) ·y (2) Derive it as However, the above processing example does not limit this exemplary embodiment.

[0066] (output unit 24) The output unit 24 outputs the prediction result derived by the prediction unit 23. As an example, the output unit 24 visually presents the prediction result to the user via a display means (not shown). The output unit 24 may also store the prediction result in a storage means (not shown). More specific processing by the output unit 24 will be described later.

[0067] (Processing flow in information processing device 1A) The flow of processing in the information processing device 1A will be described below with reference to FIG.

[0068] (Step S21) First, in step S21, the first acquisition unit 21 acquires inference data PD and a type PDC of the inference data PD. Here, the example shown in Fig. 6 illustrates a data structure in which a type PDC is included as part of the inference data PD, but this does not limit the present exemplary embodiment. Furthermore, the inference data PD acquired in this step includes one or more feature PDFs.

[0069] (Step S22-1) Subsequently, in step S22-1, the second acquisition unit 22 executes a similarity calculation process and a feature amount difference calculation process by referring to the inference data PD and the candidate data group CDG. In the example shown in FIG. 6, in this step, the following candidate data CDG is used: Candidate data CD1 whose type is C1 and contains one or more features F1 Candidate data CD2 whose type is C2 and includes one or more feature values ​​F2 An example of the similarity calculation process and feature amount difference calculation process is as follows.

[0070] (Similarity calculation process) In the similarity calculation process, the second acquisition unit 22 calculates the similarity between the inference data PD and each candidate data CD included in the candidate data group CDG as follows: The difference between the index indicating the type of the inference data PD and the index indicating the type of each candidate data CD included in the candidate data group CDG As an example, if the type is related to weather, the indicator indicating the type is calculated as follows: Sunny: 5, Sunny: 4, Cloudy: 3, Light rain: 2, Heavy rain: 1 The type of the inference data PD is "clear (index: 5)" and the types of the candidate data CD1 and CD2 are as follows: Candidate data CD1 type C1: Sunny (index: 4) Candidate data CD1 type C2: Heavy rain (index: 1) If The similarity between the inference data PD and the candidate data CD1 is calculated as 5-4=1. The similarity between the inference data PD and the candidate data CD2 is calculated as 5-1=4.

[0071] (Feature difference calculation process) In the feature amount difference calculation process, the second acquisition unit 22 The difference between one or more feature values ​​included in the inference data PD and one or more feature values ​​included in each candidate data CD included in the candidate data group CDG As an example, when the feature amount is related to the day of the week, the feature amount (which may be expressed as an index indicating the feature amount) is calculated as follows: Monday: 1, Tuesday: 2, Wednesday: 3, Thursday: 4, Friday: 5, Saturday: 6, Sunday: 7 The feature value included in the inference data PD is "Sunday (index: 7)", and the feature values ​​included in the candidate data CD1 and CD2 are Feature F1 of candidate data CD1: Saturday (index: 6) Feature F2 of candidate data CD2: Tuesday (index: 2) If Calculate the feature difference between the inference data PD and the candidate data CD1 (7-6=1), The feature difference between the inference data PD and the candidate data CD2 is calculated as 7-2=5.

[0072] The similarity and feature amount difference calculated in this step are referenced in step S22-2, which will be described later.

[0073] (Step S22-2) In step S22-2, the second acquisition unit 22 A process of comparing the similarity calculated in step S22-1 with a similarity threshold; and A process of comparing the feature amount difference calculated in step S22-1 with a threshold value for the feature amount difference. Here, the threshold value of the similarity and the threshold value of the feature amount difference may be stored in the storage unit 15A as an example, but this does not limit the present exemplary embodiment. The second acquisition unit 22 refers to the result of the comparison process and selects (identifies, determines) similar data RD that is similar to the inference data PD acquired in step S21 from the plurality of candidate data CD included in the candidate data group CDG.

[0074] For example, when the threshold value of the similarity is 2, the threshold value of the feature amount difference is 3, and the inference data PD, the candidate data CD1, and the candidate data CD2 are as described above, the second acquisition unit 22 Determine that the similarity between the inference data PD and the candidate data CD1 (1) is equal to or less than the threshold value of similarity (2); The feature difference 1 between the inference data PD and the candidate data CD1 is determined to be equal to or less than the feature difference threshold value 3, Based on the above determination, the candidate data CD1 is determined to be similar to the inference data PD.

[0075] Similarly, the second acquisition unit 22 The similarity between the inference data PD and the candidate data CD2, which is 4, is determined to be greater than the similarity threshold value, which is 2; It is determined that the feature difference between the inference data PD and the candidate data CD2, which is 5, is greater than the feature difference threshold, which is 3; Based on the above determination, the second acquiring unit 22 determines that the candidate data CD2 is not similar to the inference data PD. Based on the above determination results, the second acquiring unit 22 identifies the candidate data CD1 as similar data RD that is similar to the inference data PD.

[0076] Then, the second obtaining unit 22 searches and obtains the ratio parameters and regression coefficients associated with the similar data RD from, for example, the learning result LR stored in the storage unit 15A.

[0077] Note that the similarity and feature difference in step S22-1 and the determination process in step S22-2 are not limited to the above example. For example, in the above example, an index that becomes smaller as the two data are more similar is used as the similarity index, but conversely, an index that becomes larger as the two data are more similar may be used as the similarity index. As an example, the inverse of the similarity index described above may be used as the similarity index.

[0078] Also, in step S22-2, The product of the similarity and the feature difference may be compared with a predetermined threshold to determine whether the two data are similar or not. The product of the inverse of the similarity and the feature difference may be compared with a predetermined threshold to determine whether two pieces of data are similar.

[0079] (Step S23) Subsequently, in step S23, the prediction unit 23 derives a prediction result PR by referring to the inference data PD and the ratio parameters and regression coefficients associated with the similar data RD. Process 231: A process of applying a regression coefficient associated with the similar data RD, which is a regression coefficient for each of a plurality of target models, to the inference data PD. Process 232: Applying the ratio parameter associated with the similar data RD to the inference data PD as a ratio parameter that defines the ratio of the plurality of target models. The prediction result is derived by performing the process described above. The specific process performed by the prediction unit 23 has been described above, so a description thereof will be omitted here.

[0080] (Step S24-1) In step S24-1, the output unit 24 displays the prediction result derived by the prediction unit 23 via the input / output unit 17A. Fig. 7 shows an example of display by the output unit 24. In the example shown in Fig. 7, For the feature x0 included in the inference data PD, a regression coefficient W (a regression coefficient of the first target model) for a first target model among the plurality of target models (1) The first predicted value y calculated by applying (multiplying) (1) =W (1) ·x0 (predicted value P1 on the regression line L1 in Figure 7), For the feature x0 included in the inference data PD, a regression coefficient W for a second target model among the plurality of target models (a regression coefficient of the second target model) (2) The second predicted value y calculated by applying (multiplying) (2) =W (2) x0 (predicted value P2 on the regression line L2 in Figure 7), and a ratio parameter μ associated with said similar data RD (1) and μ (2) and the first predicted value y (1) and the second predicted value y (2) and a weighted linear sum μ (1) ·y (1) +μ (2) ·y (2) (Predicted value PP in Figure 7) is shown.

[0081] To explain with a more specific example, let us assume that each of the multiple candidate data CDs is the number of bottles of drinking water sold each day at a certain store. The candidate data CDA indicates the number of bottles of the beverage sold at the store in a clear and sunny weather (situation A). The candidate data indicating the number of bottles of the beverage sold at the store in the cloudy weather (situation B) is called the candidate data CDB. The candidate data indicating the number of bottles of the drinking water sold at the store in the rainy weather situation (situation C) is called candidate data CDC.

[0082] Examples of situations include, but are not limited to, a situation in which a campaign for the drinking water is being conducted, a situation in which the volume of the drinking water has been increased, a situation in which the appearance of the drinking water container has been changed, a situation in which a new store has opened near the store, a situation in which it is raining, and a situation in which all of these situations are included.

[0083] For example, when predicting sales of drinking water at a store when the weather is fine at a certain date and time x0, the second acquisition unit 22 acquires regression coefficients and ratio parameters obtained by a learning process that references candidate data CDA associated with "clear (situation A)" that is similar to the situation of "sunny." Then, the prediction unit 23 applies the regression coefficients and ratio parameters to the certain date and time x0 to calculate the first predicted value P1, the second predicted value P2, and the predicted value PP, and displays them as shown in FIG.

[0084] More specifically, the prediction unit 23 calculates the regression coefficient W of the model 1 obtained by the learning process with reference to the candidate data CDA. (1) , and the regression coefficient W of Model 2 (2) The first predicted value P1 and the second predicted value P2 are derived by applying each of these to a certain date and time x0. Then, the ratio parameter μ (1) and the ratio parameter μ of Model 2 (2) The predicted value PP is calculated by using the following equation.

[0085] Alternatively, the second acquisition unit 22 may acquire the regression coefficients and ratio parameters obtained by a learning process that references the candidate data CDB associated with "cloudy (situation B)" as a situation similar to the situation "sunny." Then, the prediction unit 23 may apply the regression coefficients and ratio parameters to a certain date and time x0 to calculate the first predicted value P1, the second predicted value P2, and the predicted value PP, and display them as shown in FIG.

[0086] Furthermore, the second acquisition unit 22 may acquire regression coefficients and ratio parameters obtained by a learning process with reference to candidate data CDA associated with "clear (situation A)" as a situation similar to the situation "clear," and regression coefficients and ratio parameters obtained by a learning process with reference to candidate data CDB associated with "cloudy (situation B)."The prediction unit 23 may then be configured to derive a prediction result by applying these regression coefficients and ratio parameters to a certain date and time x0.

[0087] The output unit 24 may be configured to present to the user, for each of the regression lines L1 and L2, a situation (type) associated with the regression line. More specifically, the regression coefficient W of the regression line L1 (1) The situation associated with Model 1 with the regression coefficient W of the regression line L2 (2) The configuration may further display a situation associated with the model 2 having the above.

[0088] In this way, in this step, the output unit 24 displays graphs (regression lines L1 and L2 in the above example) defined by the regression coefficients for at least two of the multiple target models (model 1 and model 2 in the above example) in a manner that allows the user to easily recognize the prediction results by the prediction unit 23.

[0089] Moreover, as described above, the information processing device 1A according to this exemplary embodiment can preferably generate the prediction result PR by referring to the type of the inference data PD. Moreover, the information processing device 1A according to this exemplary embodiment can present the prediction result PR having preferable explainability to the user.

[0090] In addition, the prediction unit 23 or the output unit 15 (24) may be configured to calculate an index (reliability index) regarding the degree of reliability of the prediction result PR by referring to the similarity and feature difference calculated in step S22-1, and display the reliability index together with the prediction result PR.

[0091] (Step S24-2) On the other hand, in the above-mentioned step S22-2, there may be a case where similar data RD similar to the inference data PD is not found (does not exist). In such a case, in this step S24-2, the output unit 24 outputs information to the effect that "a prediction result cannot be derived for the inference data PD" via the input / output unit 17A. Figure 8 shows an example of the display in this step.

[0092] (Other display examples) Note that display examples by the information processing device 1A are not limited to the above examples. As shown in FIG. 9, the output unit 24 may display a graph defined by the regression coefficients R C of each of a plurality of models. FIG. 9 shows another example of a graph displayed by the output unit 24 via the input / output unit 17A. The output unit 24 displays a graph L1 defined by the regression coefficients R C of model 1 corresponding to situation A, a graph L2 defined by the regression coefficients R C of model 2 corresponding to situation B, and a graph L3 defined by the regression coefficients R C of model 3 corresponding to situation C in a manner that allows them to be distinguished from one another. Situation C may include at least one of situations A and B, or may be a situation different from situations A and B. According to the information processing device 1A according to this exemplary embodiment, even with the above-described configuration, it is possible to present a prediction result PR having favorable explainability to the user.

[0093] Third Exemplary Embodiment A second exemplary embodiment, which is one example of an embodiment of the present invention, will be described in detail with reference to the drawings. Components having the same functions as those described in the above exemplary embodiment will be assigned the same reference numerals, and their description will be omitted as appropriate. The scope of application of each technical means employed in this exemplary embodiment is not limited to this exemplary embodiment. That is, each technical means employed in this exemplary embodiment can also be employed in other exemplary embodiments included in the present disclosure, to the extent that no particular technical obstacles arise. Furthermore, each technical means shown in each drawing referenced to explain this exemplary embodiment can also be employed in other exemplary embodiments included in the present disclosure, to the extent that no particular technical obstacles arise.

[0094] The positioning of the algorithm of the processing by the information processing device 2A according to this exemplary embodiment will be described. The present inventors have been studying a Linear Parameter-Varying (LPV) model as a modeling of a system with fluctuations. In the LPV model, as an example, an internal state quantity (internal state variable) x k , and the output state quantity (output state variable) y k is updated and calculated using the following formulas (1A) and (1B).

[0095]

number

[0096] where A (i) , B (i) is a matrix representing each state-space model (also called each end point model) that is distinguished from each other by index i, and μ (i) k is a parameter that defines the internal division ratio (weight) of each model. μ (i) k These are called internal ratio parameters, weight parameters, or scheduling parameters. In the above LPV model, u kare input variables, and C and D are x k and u k k is an index assigned to each state variable, and is, for example, time.

[0097] Figure 11 shows the output of each end point model in the LPV model (1st SS model to 5th SS model in Figure 11) and the internal division ratio parameter μ (i) k As shown in FIG. 11, the output of multiple endpoint models at the k-th step is (A (i) x k +B (i) u k ) (i=1~5) For each of the internal division ratio parameters μ (i) k (i=1~5) is multiplied, and x at the k+1th step is k+1 is calculated.

[0098] While such an LPV model is suitable for modeling systems with fluctuations, the internal ratio parameter μ (i) k However, there was a problem that it was difficult to apply to systems where the value of is unknown.

[0099] The present inventors The above internal division ratio parameter μ (i) k The hidden variable (posterior probability) z k Treat it as - Applying hidden variable model learning methods in machine learning, The above internal division ratio parameter μ (i) k the hidden variable z k Calculate as the expected value of By doing so, the inventors have found that it is possible to realize the learning of an LPV model even if the interior ratio parameter μ (i)k The L2PV model (Latent Linear Parameter-Varying model) is defined by the following equations (2A) to (2C), which introduce the following as hidden variables:

[0100]

number

[0101] The following equations (3A) to (3E) are used:

[0102]

number

[0103] By rewriting it into the regression model form (L2PV regression model) specified by (i) k This gave rise to the idea of ​​using this as a learning subject.

[0104] Each process performed by the information processing device 2A described below is based on the above-mentioned formulation and is a process based on the unique viewpoint of the inventor.

[0105] (Configuration of information processing device 2A) The configuration of the information processing device 2A will be described with reference to Fig. 10. Fig. 10 is a block diagram showing the configuration of the information processing device 2A. As shown in Fig. 10, the information processing device 2A includes a control unit 20A, a storage unit 25A, a communication unit 26A, and an input / output unit 27A.

[0106] (Storage section 25A) First, various data (information) stored in the storage unit 25A will be described. Data referenced by the control unit 20A is stored in the storage unit 25A. Examples of the storage unit 25A include, but are not limited to, a flash memory, a hard disk drive (HDD), a solid state drive (SSD), or a combination thereof.

[0107] Examples of data stored in memory unit 25A include, but are not limited to, target data TD, ratio parameter RP, regression coefficient RC, type information CI, distribution information DI, learning result LR, inference data PD, and prediction result PR, as shown in FIG. 10.

[0108] (Explanation of target data TD (Part 1)) The target data TD is data used in the learning process in the information processing device 2A. The target data TD is a state variable (~x k ) and state variables (~y k ) is expressed as the following equation (4). k ,~x k ,y k ,~y k are sometimes called features. Also, the state variable x k ,~x k are called explanatory variables, and the state variable y k ,~y k The objective variable is sometimes called a target variable. Furthermore, when the objective variable is the target of derivation, the objective variable is sometimes called a predicted value. These specific names do not limit the contents described in this specification.

[0109]

number

[0110] The ratio parameter RP is a parameter that defines the relative weights of a plurality of state space models in an LPV model, and is also referred to as a scheduling parameter. The ratio parameter RP may be an internal ratio parameter or an external ratio parameter. In this exemplary embodiment, in order to describe the case where the ratio parameter RP is an internal ratio parameter, the ratio parameter RP is also referred to as an internal ratio parameter RP. The ratio parameter RP is also referred to as a weight parameter RP or a scheduling parameter RP. As an example, the ratio parameter RP is given by the following equation (5) corresponding to each of m models (model 1 to model m):

[0111]

number

[0112] Here, k is an index similar to the index assigned to each state variable described above, and N represents the dimension of each state variable (the number of samples of each state variable). Also, the index (i) relating to the model is not explicitly stated in the above expression. This is because the ratio parameter RP is divided into an internal ratio parameter vector consisting of components corresponding to models 1 to m for each k. μ k =(μ k (1) , μ k (2) , , μ k (m) ) In this way, the ratio parameter RP may be expressed as an internal division ratio parameter vector or a ratio parameter matrix.

[0113] Also, the internal ratio parameter μ for a certain model j k (j) is the N-dimensional target data x k It can also be expressed as the components of an N-dimensional vector having components corresponding to each of the j-th internal division ratio parameters μ k (j)is the N-dimensional target data x k (k=1 to N) (j) , μ2 (j) , , μ N (j) ) are the components of an N-dimensional vector with

[0114] The storage unit 25A may be configured to store multiple pieces of target data TD. In this case, the multiple pieces of target data TD may be referred to as a target data group TDG. Furthermore, since the target data TD also serves as candidate data referenced in the prediction process by the prediction unit 18 (23), which will be described later, the target data TD may be referred to as candidate data CD, and the target data group TDG may be referred to as a candidate data group CDG. However, these names do not limit this exemplary embodiment.

[0115] The regression coefficient RC is a coefficient in the L2PV regression model. The regression coefficient RC is expressed as the following equation (6).

[0116]

number

[0117] (Type Information CI) The type information CI is information indicating the type of the inference data PD, which will be described later. Here, the information indicating the type may be, for example, information indicating the circumstances under which the inference data PD was acquired. The type information CI may also be configured to include information regarding the type of the target data TD (candidate data CD).

[0118] (Explanation of target data TD (part 2)) As described above, the target data TD (target data group TDG) is also positioned as candidate data CD (candidate data CDG). Below, the target data TD will be described from the perspective of candidate data CD.

[0119] The candidate data group CDG includes one or more candidate data CDs that are referenced in the inference process executed by the information processing device 2A. Here, each candidate data CD is, for example, Information about the type of candidate data CD One or more features Here, the information on the type of the candidate data CD may be, for example, information indicating the circumstances under which each candidate data CD was acquired. For example, when the one or more feature quantities include "day" as an explanatory variable, the information on the type may be, Candidate data CD1 (e.g., daily sales volume): Situation A (e.g., sunny weather) Candidate data CD2 (e.g., daily sales volume): Situation B (e.g., cloudy weather) Candidate data CD3 (e.g., daily sales volume): Situation C (e.g., the weather is rainy) The information may be associated with each candidate data CD as follows.

[0120] In addition, information about the above types is available at Information indicating whether a certain candidate data is similar to other candidate data or data for inference Here, "a certain piece of data is similar to another piece of data" refers to, for example, a case where the type of the certain piece of data is similar to the type of the other piece of data. For example, if candidate data 1 is data on the number of sales acquired in situation A (e.g., sunny weather) as in the above example, candidate data or data for inference that is data on the number of sales in situation A' (e.g., sunny weather) similar to situation A is data similar to candidate data 1.

[0121] The storage unit 25A may be configured to further store correspondence relationship information, which will be described below, as part of the information about the type, or together with the information about the type. The correspondence relationship information is information about the correspondence relationship between each candidate data CD and a plurality of target models. Examples of the information about the correspondence relationship include: - Information indicating what kind of candidate data each of the multiple target models CD has been trained (will be trained) with is included. For example, the correspondence information regarding the target models a, b, and c and the candidate data CD1, CD2, and CD3 is as follows: Target model a: Candidate data CD1, CD2 Target model b: Candidate data CD3 Target model c: Candidate data CD1, CD2, CD3 Here, the first line in the above example indicates that the target model a has been trained (will be trained) using the candidate data CD1 and the candidate data CD2. The same applies to the other lines.

[0122] In other words, the correspondence information is Information indicating whether each of the multiple target models has used (will use) each of the multiple candidate data CD as training data As an example, the correspondence relationship information CI regarding the target models a, b, and c and the candidate data CD1, CD2, and CD3 may be expressed as follows: Target model a: Candidate data CD1 (○), CD2 (○), CD3 (×) Target model b: Candidate data CD1(×), CD2(×), CD3(○) Target model c: Candidate data CD1(○), CD2(○), CD3(○) Here, ○ indicates that it was used (will be used) in learning, and × indicates that it was not used (will not be used) in learning.

[0123] In addition, when the correspondence information includes the correspondence between each candidate data CD and the situation in which the candidate data CD was acquired as described above, the correspondence information may be expressed as including the relationship between each target model and the type related to the target model. For example, in the above example, the correspondence information is Target model a: Situation A (e.g., sunny weather), Situation B (e.g., cloudy weather) Target model b: Situation C (e.g., rainy weather) Target model c: Situations A, B, and C The correspondence information may include information indicating a correspondence such as the following. By including the correspondence information as described above, the information processing device 2A can identify what situation each target model is associated with, in other words, what situation each target model is preferably used in. Furthermore, the information processing device 2A can present to the user what situation each target model is preferably used in.

[0124] The distribution information DI includes a covariance matrix Φ of the prior distribution of the latent variables, a covariance parameter η of the prior distribution of the latent variables, and a covariance parameter Ψ of the posterior distribution of the latent variables.

[0125] Hidden variable z k The prior distribution p(z k ) is expressed as the following equation (7).

[0126]

number

[0127] Hidden variable z k In other words, the covariance parameter η of the prior distribution of k |z k ,~x k ,W,η) is the covariance parameter η.

[0128]

number

[0129] Hidden variable z k The posterior distribution p(z k |~y k ,~x k ,W,η) is expressed as the following equation (9).

[0130]

number

[0131] In the above formula, the calligraphy font N on the right side represents a normal distribution. However, this does not mean that the example of the distribution in this exemplary embodiment is limited to a normal distribution. As an example, the hidden variable z k The Dirichlet distribution may be used as the posterior distribution of .

[0132] As will be described later, in the processing by the information processing device 2A, the hidden variable z k The posterior distribution p(z k |~y k ,~x k , W, η) is expressed under the constraints (constraints) of the following equation (10).

[0133]

number

[0134] Therefore, the hidden variable z k Even when a normal distribution is used as the posterior distribution of , a suitable calculation can be performed.

[0135] The learning result LR is data that is output by the output unit 15, which will be described later. The learning result LR includes the calculated ratio parameter RP and the calculated regression coefficient RC.

[0136] The inference data PD is an internal state quantity input to the L2PV regression model. The L2PV regression model makes predictions by using the inference data PD as input and applying the calculated regression coefficients RC to each of multiple target models.

[0137] The predicted result PR is a predicted result obtained by the L2PV regression model. Examples of the predicted result PR will be described later.

[0138] (Communication unit 26A) The communication unit 26A is an interface for transmitting and receiving data via a network. Examples of the communication unit 26A include, but are not limited to, communication chips for various communication standards such as Ethernet (registered trademark), Wi-Fi (Wireless Fidelity (registered trademark), and wireless communication standards for mobile data communication networks, and USB-compliant connectors.

[0139] (Input / output section 27A) The input / output unit 27A is an interface that receives input of data and outputs data. Examples of the input / output unit 27A include, but are not limited to, a microphone, a camera, a gaze input device, a keyboard, a touchpad, a speaker, and a liquid crystal display.

[0140] (Control unit 20A) The control unit 20A controls each component included in the information processing device 2A. As shown in FIG. 10 , the control unit 20A includes an acquisition unit 11, a regression coefficient calculation unit 12, a covariance calculation unit 13, a ratio parameter calculation unit 14, an output unit 15, an initial value determination unit 16, a convergence determination unit 17, and a prediction unit 18. In this exemplary embodiment, the acquisition unit 11, the regression coefficient calculation unit 12, the covariance calculation unit 13, the ratio parameter calculation unit 14, the initial value determination unit 16, the convergence determination unit 17, and the prediction unit 18 function as an acquisition unit, a regression coefficient calculation unit, a covariance calculation unit, a ratio parameter calculation unit, an initial value determination unit, a convergence determination unit, and a prediction unit, respectively. In this exemplary embodiment, the output unit 15 functions as a storage unit and an output unit. Specific examples of the processing performed by each unit will be described later with reference to different drawings.

[0141] Note that the acquisition unit 11 also functions as a first acquisition unit 21 and a second acquisition unit 22 included in the information processing device 1A according to the second exemplary embodiment, and therefore the acquisition unit 11 may be referred to as an acquisition unit 11 (21, 22). Furthermore, the output unit 15 also functions as an output unit 24 included in the information processing device 1A, and therefore the output unit 15 may be referred to as an output unit 15 (24). Furthermore, the prediction unit 18 also functions as a prediction unit 23 included in the information processing device 1A, and therefore the prediction unit 18 may be referred to as a prediction unit 18 (23).

[0142] The acquisition unit 11 (21, 22) acquires data via the communication unit 26A or the input / output unit 27A. Examples of data acquired by the acquisition unit 11 (21, 22) include information on the target data TD, the number of models of multiple target models, and the type of the target data TD. Another example of data acquired by the acquisition unit 11 is information on the latent variable z k , information on the prior distribution of the data PD. Still another example of data acquired by the acquisition unit 11 is information on the inference data PD and the type of the inference data. The acquisition unit 11 stores the acquired data in the storage unit 15A.

[0143] The regression coefficient calculation unit 12 calculates a regression coefficient RC for each of the plurality of target models by referring to a ratio parameter RP that defines the ratio of the plurality of target models and the target data TD. As one example, the ratio parameter RP referred to by the regression coefficient calculation unit 12 is the initial value of the ratio parameter RP determined by an initial value determination unit 16, which will be described later. As another example, the ratio parameter RP referred to by the regression coefficient calculation unit 12 is the ratio parameter RP calculated by a ratio parameter calculation unit 14, which will be described later. The regression coefficient calculation unit 12 stores the calculated regression coefficient RC in the storage unit 15A.

[0144] The covariance calculation unit 13 calculates the target data TD, the ratio parameter RP, the regression coefficient RC, and the latent variable z k By referring to the covariance matrix Φ of the prior distribution of the hidden variable z k The covariance parameter η of the prior distribution of and the hidden variable z kThe covariance calculation unit 13 calculates the covariance matrix Ψ of the posterior distribution of the calculated hidden variable z k The covariance parameter η of the prior distribution of and the hidden variable z k and the covariance matrix Ψ of the posterior distribution of the above are stored in the storage unit 15A as distribution information DI.

[0145] The ratio parameter calculation unit 14 calculates the ratio parameter RP by referring to the target data TD and the regression coefficient RC. k The ratio parameter calculation unit 14 further refers to the covariance matrix Ψ of the posterior distribution of the ratio parameter RP to calculate the ratio parameter RP. The ratio parameter calculation unit 14 stores the calculated ratio parameter RP in the storage unit 25A.

[0146] Furthermore, the ratio parameter calculation unit 14 refers to the correspondence information CI and calculates the ratio parameter RP under a constraint condition according to the correspondence. As an example, the ratio parameter calculation unit 14 calculates the ratio parameter RP under a constraint condition on the sum of the ratio parameters RP according to the correspondence.

[0147] Furthermore, the ratio parameter calculation unit 14 calculates the ratio parameter RP using an update equation that includes a normalization term according to the correspondence relationship.

[0148] The output unit 15 (24) stores the regression coefficient RC and the ratio parameter RP in association with the target data TD. As an example, the output unit 15 (24) supplies the regression coefficient RC and the ratio parameter RP to the storage unit 25A in association with information on the type of the target data TD.

[0149] Furthermore, the output unit 15 (24) outputs the regression coefficient RC and the ratio parameter RP (learning result LR) calculated by the ratio parameter calculation unit 14. As one example, the output unit 15 (24) outputs an image including the regression coefficient RC and the ratio parameter RP calculated by the ratio parameter calculation unit 14 to the input / output unit 27A. In this configuration, the output unit 15 (24) may display graphs defined by the regression coefficients RC for at least two of the multiple target models in a manner that allows them to be distinguished from one another. As another example, the output unit 15 (24) outputs the regression coefficient RC and the ratio parameter RP calculated by the ratio parameter calculation unit 14 when the convergence determination unit 17 (described later) determines that the calculation related to the ratio parameter RP has converged.

[0150] The initial value determination unit 16 determines the initial value of the ratio parameter RP to be referred to by the regression coefficient calculation unit 12. The initial value determination unit 16 stores the determined initial value of the ratio parameter RP in the storage unit 25A.

[0151] The convergence determination unit 17 determines whether or not the calculation regarding the ratio parameter RP has converged, and supplies the determination result to the output unit 15 (24).

[0152] The prediction unit 18 (23) derives a prediction result PR by applying, to the inference data PD, the regression coefficients RC for each of the multiple target models and the ratio parameters RP calculated by the ratio parameter calculation unit 14, which are determined according to the type of the inference data PD. The prediction unit 18 (23) stores the derived prediction result PR in the storage unit 15A. Note that, since more specific processing by the prediction unit 18 (23) is similar to that of the prediction unit 23 according to the second exemplary embodiment, a description thereof will be omitted here.

[0153] (Example of processing flow in information processing device 2A) 12 is a diagram showing an example of a processing flow in an information processing device 2A according to this exemplary embodiment. Note that the processing example described below can also be regarded as a variational Bayes EM algorithm, but this does not limit this exemplary embodiment. Furthermore, the processing example described below can be regarded as processing for updating each parameter so as to maximize a variational lower bound (VLB) J obtained by the following equation (11).

[0154]

number

[0155] The processing example described below can also be expressed as an algorithm for solving a maximum likelihood problem defined by the model likelihood p in the following equation (12).

[0156]

number

[0157] (Step S11: Acquisition process) In step S11, the acquisition unit 11 (21, 22) acquires target data TD. Here, as described above, the target data TD is data used in the learning process in the information processing device 2A, and is also positioned as candidate data CD. Details of the target data TD have been explained, so explanation will be omitted here.

[0158] In step S11, the acquisition unit 11 (21, 22) further acquires a parameter m indicating the number of models among the plurality of target models. Here, the number of models m is determined by the internal division ratio parameter vector μ k (i) It can also be expressed as the number of

[0159] Also, in step S11, the acquisition unit 11 (21, 22) acquires information about the type of the target data TD.

[0160] In step S11, the acquisition unit 11 (21, 22) acquires the hidden variable z k As an example, the acquisition unit 11 (21, 22) acquires information about the prior distribution of the hidden variable z k The prior distribution p(z k ) covariance matrix Φ. The acquisition unit 11 (21, 22) also acquires the covariance matrix Φ of the hidden variable z k The prior distribution p(z k ) may further be obtained.

[0161] (Step S16: Initial value determination process) Next, in step S16, the initial value determination unit 16 determines the initial value of a ratio parameter RP to be referenced in the regression coefficient calculation process S12, which will be described later. As an example, the initial value determination unit 16 determines the initial value of the ratio parameter RP as a random value. By the initial value determination unit 16 determining the initial value of the ratio parameter RP in this way, the regression coefficient RC can be suitably calculated in the regression coefficient calculation process S12, which will be described later. The details of the ratio parameter RP have been explained above, so a detailed explanation will be omitted here.

[0162] (Step S12: Regression coefficient calculation process) Subsequently, in step S12, the regression coefficient calculation unit 12 calculates the ratio parameter (internal division ratio parameter vector) RP expressed as the following equation (13):

[0163]

number

[0164] and the target data TD expressed as the following equation (14):

[0165]

number

[0166] and a regression coefficient RC for each of the plurality of target models, which is expressed as the following equation (15) with reference to

[0167]

number

[0168] As an example, the regression coefficient calculation unit 12 calculates a regression coefficient RC expressed as the following equation (17) by referring to the ratio parameter RP and the target data TD using the following equation (16).

[0169]

number

[0170]

number

[0171] Here, the asterisk on the shoulder of W indicates the updated value, and in the calculation formula, the operation symbol with a cross in a circle represents the Kronecker product. T represents transposition. Also, Ψ k is the hidden variable z k represents the covariance parameter of the posterior distribution of

[0172] (Step S13: Covariance calculation process) Subsequently, in step S13, the covariance calculation unit 13 calculates the target data TD expressed as the following equation (18):

[0173]

number

[0174] and the ratio parameter (internal division ratio parameter vector) RP expressed as the following equation (19):

[0175]

number

[0176] and the regression coefficient RC expressed as the following equation (20):

[0177]

number

[0178] and the hidden variable z k The hidden variables z k The covariance parameter η of the prior distribution of the hidden variable z k The covariance matrix of the posterior distribution of {Ψ k} k=1 N As an example, the covariance calculation unit 13 calculates the following equation (21):

[0179]

number

[0180] (where Λ k is given by the following equation (22)

[0181]

number

[0182] By the hidden variable z k The covariance matrix of the posterior distribution of {Ψ k} k=1 N Furthermore, the covariance calculation unit 13 calculates the following equation (23):

[0183]

number

[0184] By the hidden variable z k Calculate the covariance parameter η of the prior distribution of , where N is the number of samples for each state variable as described above, and r is the covariance parameter η of the prior distribution of . k , and as an example, r=1.

[0185] (Step S14: Ratio parameter calculation process) Subsequently, in step S14, the ratio parameter calculation unit 14 calculates the target data TD expressed as the following equation (24):

[0186]

number

[0187] and the regression coefficient RC expressed as the following equation (25):

[0188]

number

[0189] and the hidden variable z k The covariance matrix of the posterior distribution of {Ψ k} k=1 N With reference to the above, the ratio parameter (internal division ratio parameter vector) RP expressed as the following equation (26) is

[0190]

number

[0191] As an example, the ratio parameter calculation unit 14 calculates (updates) the following equation (27):

[0192]

number

[0193] By μ k The ratio parameter RP may be calculated (updated) by executing the process of calculating the ratio parameter RP under constraints according to at least one of the type information CI and the correspondence indicated by the correspondence information.

[0194]

number

[0195] By μ k By performing the process of calculating the ratio parameter (internal division ratio parameter vector) RP expressed by the following equation (30) under the constraint conditions (constraint conditions) expressed by the following equations (29A) to (29C),

[0196]

number

[0197]

number

[0198] Here, the first equation of the constraint condition (29A) can be expressed by explicitly specifying the index (i) related to the model as follows: Σ i=1 m μ k (i) = 1 In other words, the first equation of the constraint condition equation (29A) indicates that the sum of the ratio parameter RP across the indexes related to the model is 1. Furthermore, the second equation of the constraint condition equation (29A) indicates that the value of the ratio parameter RP is equal to or greater than 0. In this way, by calculating the ratio parameter RP under the constraint condition equation (29A), the ratio parameter calculation unit 14 can suitably calculate the ratio parameter RP even when, for example, a normal distribution is adopted as the posterior distribution of the latent variable.

[0199] In addition, in the above constraint equation (29B), M i denotes the set of target data TD corresponding to the i-th endpoint model. In other words, M iindicates a set of target data TD whose correspondence information (type information CI) indicates that it is to be used as training data for the i-th end point model. Therefore, the above constraint condition formula (29B) indicates that the ratio parameter μ calculated using the j-th target data TD from the set of target data TD that indicates that it is to be used as training data for the i-th end point model. j The sum of these is 1. That is, by using the above constraint condition formula (29B), the ratio parameter calculation unit 14 calculates the ratio parameter under the constraint condition on the sum of the ratio parameters RP, which is a constraint condition according to the correspondence (correspondence information, type information CI). On the other hand, the above constraint condition formula (29C) shows that the ratio parameter μ in the endpoint model that did not use the j-th target data TD as training data h indicates that ∇ ...

[0200] For example, consider a model that predicts sales of drinking water in a store. As an example, the endpoint models are (1) Model 1, which predicts sales when a measure is implemented, (2) Model 2, which predicts sales when the content volume is changed, (3) Model 3, which predicts sales when the appearance is changed, (4) Model 4, which predicts sales when the external environment changes, and Model 5, which predicts sales under all of the situations (1) to (5).

[0201] As an example of this case, if the target data TD is sales data when a measure is implemented, the correspondence information (type information CI) indicates that the target data TD will be used as learning data for Model 1 and Model 5.

[0202] As described above, the ratio parameter calculation unit 14 may calculate the ratio parameter RP using an update equation including a normalization term according to the correspondence relationship. For example, the ratio parameter calculation unit 14 calculates the ratio parameter RP using the following equations (31A) and (31B).

[0203]

number

[0204] Of the equation (31A), the following equation (32) is a normalization term according to the correspondence relationship (type information CI).

[0205]

number

[0206] In equation (31A), γ k is a coefficient indicating the strength of regularization. That is, equation (32) is the ratio parameter μ calculated using the k-th target data TD. k is the ratio parameter μ calculated using target data TD similar to the k-th target data TD k’ This indicates that the value is close to . With this configuration, when two pieces of target data TD are similar to each other and the ratio parameter RP is calculated using each of the two pieces of target data TD, the ratio parameter calculation unit 14 calculates ratio parameters RP that are close in value. Therefore, the ratio parameter calculation unit 14 can suitably calculate the ratio parameter RP by referring to the similarity of the target data TD. Furthermore, it is possible to calculate a ratio parameter RP with improved explainability.

[0207] (Step S17: Convergence determination process) Subsequently, in step S17, the convergence determination unit 17 determines whether or not the series of processes in the above-mentioned steps S12, S13, and S14 have converged. This may be expressed as determining whether or not the above-mentioned variational Bayes EM algorithm has converged, or as determining whether or not the calculation related to the ratio parameter RP in step S14 has converged. As an example, the convergence determination unit 17 determines whether or not the variational lower bound (VLB) J obtained by the following formula (33),

[0208]

number

[0209] and determines that the series of processes in steps S12, S13, and S14 described above have converged if the change in the variation lower limit is equal to or less than a predetermined threshold. For example, in the nth iteration of the series of processes including steps S12, S13, and S14 described above, the convergence determination unit 17 compares the (n-1)th variation lower limit with the nth variation lower limit in the convergence determination process, and determines that the series of processes in steps S12, S13, and S14 described above have converged if the absolute value of the difference between them is equal to or less than a predetermined threshold.

[0210] (Step S15: Output process) If the convergence determination unit 17 determines in step S17 that the calculation has "converged," then in step S15, the output unit 15 (24) outputs the regression coefficient RC calculated by the regression coefficient calculation unit 12 in step S12 and the ratio parameter RP (learning result LR) calculated by the ratio parameter calculation unit 14 in step S14. As an example, the output unit 15 (24) associates the regression coefficient RC and the ratio parameter RP with the type of target data TD and stores them in the storage unit 25A. The stored target data TD is then referenced as candidate data CD in the prediction process. In this way, if the convergence determination unit 17 determines that the calculation regarding the ratio parameter RP has "converged," the output unit 15 outputs the learning result LR (24), thereby making it possible to output a suitable learning result LR.

[0211] In step S15, the output unit 15 (24) outputs the hidden variable z calculated by the covariance calculation unit 13 in step S13. k The covariance parameter η of the prior distribution of the hidden variable z k With this configuration, the output unit 15 (24) may further output the covariance matrix Ψ of the posterior distribution of the hidden variable z k The covariance parameter η of the prior distribution of and the hidden variable z k The covariance matrix Ψ of the posterior distribution of σ can be presented to the user.

[0212] In step S15, the output unit 15 may be configured to display graphs defined by the regression coefficients RC for at least two of the plurality of target models in a manner that allows them to be distinguished from one another.

[0213] 13 shows an example of a graph that the output unit 15 (24) displays via the input / output unit 27A in this step. In the example shown in FIG. 13, the output unit 15 (24) displays the regression coefficients RC calculated for each of the multiple target models, which are expressed as the following equation (34), in the regression coefficient calculation process in step S12.

[0214]

number

[0215] Among them, the regression coefficient W of Model 1 (1) The graph L1 defined by the regression coefficient W of Model 2 (2) and graph L2 defined by the above are displayed so as to be distinguishable from each other. In this way, according to the information processing device 2A according to this exemplary embodiment, a plurality of models are used and the ratio parameter RP of each model can be determined by learning, so that an output result having a range (as an example, an output result having a range defined by the above graph L1 and graph L2) can be generated.

[0216] 13, the output unit 15 may refer to the correspondence relationship information CI and display what models the graph L1 and the graph L2 correspond to. In Fig. 13, the output unit 15 displays that the graph L1 is a model corresponding to the situation A, and the graph L2 is a model corresponding to the situation B.

[0217] (Example of processing executed by the prediction unit 18 (23)) The prediction unit 18 (23) derives multiple prediction results PR by applying, to the inference data PD, the regression coefficient RC for each of the multiple target models and the ratio parameter RP determined according to the type of the inference data PD, among the ratio parameters RP calculated by the ratio parameter calculation unit 14. As an example, when the inference data PD includes a feature amount (explanatory variable) x0, the prediction unit 18 (23) derives the prediction results PR by applying, to the feature amount x0, the regression coefficient RC for each of the multiple target models and the ratio parameter RP determined according to the type of the inference data PD.

[0218] The ratio parameter determined according to the type of inference data PD refers to, for example, a ratio parameter learned (updated) by the above-mentioned learning process using learning data (target data TD, candidate data CD) associated with a type that is the same as or similar to the type of the inference data PD, but this does not limit the present exemplary embodiment. Note that the more specific processing by the prediction unit 18 (23) is similar to that of the prediction unit 23 according to exemplary embodiment 2, and therefore will not be described here.

[0219] (Effects of information processing device 2A) As described above, in the information processing device 2A, Acquire information regarding the target data, the number of models of multiple target models, and the type of the target data; calculating a regression coefficient for each of the plurality of target models by referring to a ratio parameter that defines a ratio of the plurality of target models and the target data; Calculating the ratio parameter by referring to the target data and the regression coefficients; The regression coefficients and the ratio parameters are stored in association with information about the type. The information processing device 2A has the following configuration. Therefore, in an optimization method that refers to a plurality of models, information about the type is stored in association with the regression coefficients and the ratio parameters. Then, the information about the type can be suitably used in the prediction phase. Therefore, the information processing device 2A provides a technique for suitably deriving prediction results in an optimization method that refers to a plurality of models.

[0220] Fourth Exemplary Embodiment A fourth exemplary embodiment, which is an example of an embodiment of the present invention, will be described in detail with reference to the drawings. Components having the same functions as those described in the above exemplary embodiment will be assigned the same reference numerals, and their description will be omitted as appropriate. The scope of application of each technical means employed in this exemplary embodiment is not limited to this exemplary embodiment. That is, each technical means employed in this exemplary embodiment can also be employed in other exemplary embodiments included in the present disclosure, to the extent that no particular technical obstacles arise. Furthermore, each technical means shown in each drawing referenced to explain this exemplary embodiment can also be employed in other exemplary embodiments included in the present disclosure, to the extent that no particular technical obstacles arise.

[0221] (Configuration of information processing device 2B) The configuration of the information processing device 2B will be described with reference to Fig. 14. Fig. 14 is a block diagram showing the configuration of the information processing device 2B. As shown in Fig. 14, the information processing device 2B includes a control unit 20B, a storage unit 25B, a communication unit 26A, and an input / output unit 27A. The communication unit 26A and the input / output unit 27A have been described above, and therefore their description will be omitted.

[0222] Similar to the above-described storage unit 25A, the storage unit 25B stores data referenced by the control unit 20B. Examples of data stored in the storage unit 25B include, but are not limited to, target data TD, ratio parameter RP, regression coefficient RC, type information CI, learning result LR, inference data PD, and prediction result PR, as shown in Fig. 14. The target data TD, ratio parameter RP, regression coefficient RC, type information CI, learning result LR, inference data PD, and prediction result PR have been described above, and therefore further description will be omitted.

[0223] (control unit 20B) The control unit 20B controls each component included in the information processing device 2B, similar to the control unit 20A described above. As shown in FIG. 14 , the control unit 20B includes an acquisition unit 11 (21, 22), a regression coefficient calculation unit 12, a ratio parameter calculation unit 14, an output unit 15 (24), an initial value determination unit 16, a convergence determination unit 17, and a prediction unit 18 (23). In this exemplary embodiment, the acquisition unit 11 (21, 22), the regression coefficient calculation unit 12, the ratio parameter calculation unit 14, the initial value determination unit 16, the convergence determination unit 17, and the prediction unit 18 (23) respectively implement an acquisition unit, a regression coefficient calculation unit, a ratio parameter calculation unit, an initial value determination unit, a convergence determination unit, and a prediction unit. The output unit 15 (24) implements a storage unit and an output unit. The acquisition unit 11 (21, 22), output unit 15 (24), initial value determination unit 16, convergence determination unit 17, and prediction unit 18 (23) have been described in the third exemplary embodiment, and therefore description thereof will be omitted.

[0224] The regression coefficient calculation unit 12 calculates a regression coefficient RC for each of the plurality of target models by referring to the ratio parameter RP that defines the ratio of the plurality of target models and the target data TD. In this exemplary embodiment, the regression coefficient calculation unit 12 calculates the regression coefficient RC for each of the plurality of target models by referring to the hidden variable z k The covariance parameter η of the prior distribution of and the hidden variable z k The regression coefficients RC are calculated using the least squares method without updating the covariance matrices Φ and Φ of the prior distributions.

[0225] The ratio parameter calculation unit 14 calculates the ratio parameter RP by referring to the target data TD and the regression coefficient RC. In this exemplary embodiment, the ratio parameter RP is not limited to an internal ratio parameter, and may be an external ratio parameter.

[0226] (Example of processing flow in information processing device 2B) 15 is a diagram showing an example of the flow of processing in the information processing device 2B according to this exemplary embodiment. The acquisition processing S11, the initial value determination processing S16, and the output processing S15 are as described above, and therefore descriptions thereof will be omitted.

[0227] (Step S12: Regression coefficient calculation process) In step S12, the regression coefficient calculation unit 12 calculates the ratio parameter RP expressed as the following equation (35):

[0228]

number

[0229] and the target data TD expressed as the following equation (36):

[0230]

number

[0231] and the regression coefficient RC for each of the plurality of target models, which is expressed as the following equation (37):

[0232]

number

[0233] As an example, the regression coefficient calculation unit 12 calculates the regression coefficient W that minimizes the value of the evaluation function of the following equation (38) with the value of the ratio parameter RP and the value of the target data TD set as fixed values. (i) Calculate the value of

[0234]

number

[0235] That is, the regression coefficient calculation unit 12 calculates the regression coefficient W using the least squares method.

[0236] (Step S14: Ratio parameter calculation process) Subsequently, in step S14, the ratio parameter calculation unit 14 calculates the target data TD expressed as the following equation (39):

[0237]

number

[0238] and the regression coefficient RC expressed as the following equation (40):

[0239]

number

[0240] and the ratio parameter RP expressed as the following equation (41)

[0241]

number

[0242] As described above, the ratio parameter calculation unit 14 calculates (updates) the hidden variables z k The covariance parameter η of the prior distribution of and the hidden variable z k The ratio parameter RP is calculated without updating the covariance matrix Φ of the prior distribution of . -1 The ratio parameter RP is calculated by the following equation (42) with ρ = 0:

[0243]

number

[0244] Furthermore, similarly to the above-described exemplary embodiment, the ratio parameter calculation unit 14 calculates the ratio parameter RP under the constraints expressed as the following equations (43A) and (43B).

[0245]

number

[0246] (Step S17: Convergence determination process) Subsequently, in step S17, the convergence determination unit 17 determines whether or not the series of processes in steps S12, S13, and S14 described above has converged. In step S17, the convergence determination unit 17, for example, determines the variational lower bound (VLB) J obtained by the following equation (44), as in the exemplary embodiment:

[0247]

number

[0248] If the change in the variation lower limit is equal to or less than a predetermined threshold, it is determined that the series of processes in steps S12, S13, and S14 have converged.

[0249] As another example, the convergence determination unit 17 determines whether the value of the ratio parameter RP has converged. For example, in the nth convergence determination process of repeating a series of processes including the above-mentioned steps S12, S13, and S14, the convergence determination unit 17 compares the value of the ratio parameter RP at the (n-1)th time with the value of the ratio parameter RP at the nth time, and determines that the series of processes at the above-mentioned steps S12, S13, and S14 has converged if the absolute value of the difference between them is equal to or smaller than a predetermined threshold.

[0250] (Effects of information processing device 2B) As described above, the information processing device 2B calculates the regression coefficient R C using the least squares method. With this configuration, the information processing device 2B also provides a technique for suitably deriving a prediction result in an optimization method that references multiple models, similar to the information processing device 2A.

[0251] [Software implementation example] Some or all of the functions of the information processing devices 1, 1A, 2, 2A, and 2B (hereinafter also referred to as "each of the above devices") may be realized by hardware such as an integrated circuit (IC chip), or by software.

[0252] In the latter case, each of the above devices is realized by, for example, a computer that executes instructions of a program, which is software that realizes each function. An example of such a computer (hereinafter referred to as computer C) is shown in Figure 15. Figure 15 is a block diagram showing the hardware configuration of computer C that functions as each of the above devices.

[0253] The computer C includes at least one processor C1 and at least one memory C2. The memory C2 stores a program P for causing the computer C to operate as each of the above-mentioned devices. In the computer C, the processor C1 reads and executes the program P from the memory C2, thereby realizing the functions of each of the above-mentioned devices.

[0254] The processor C1 may be, for example, a central processing unit (CPU), a graphic processing unit (GPU), a digital signal processor (DSP), a micro processing unit (MPU), a floating point number processing unit (FPU), a physics processing unit (PPU), a tensor processing unit (TPU), a quantum processor, a microcontroller, or a combination thereof. The memory C2 may be, for example, a flash memory, a hard disk drive (HDD), a solid state drive (SSD), or a combination thereof.

[0255] The computer C may further include a RAM (Random Access Memory) for expanding the program P during execution and for temporarily storing various data. The computer C may also include a communication interface for transmitting and receiving data to and from other devices. The computer C may also include an input / output interface for connecting input / output devices such as a keyboard, mouse, display, and printer.

[0256] Furthermore, the program P can be recorded on a non-transitory tangible recording medium M that can be read by the computer C. Such a recording medium M can be, for example, a tape, a disk, a card, a semiconductor memory, or a programmable logic circuit. The computer C can acquire the program P via such a recording medium M. The program P can also be transmitted via a transmission medium. Such a transmission medium can be, for example, a communication network or broadcast waves. The computer C can also acquire the program P via such a transmission medium.

[0257] [Appendix A] This disclosure includes the techniques described in the following appendices. However, the present invention is not limited to the techniques described in the following appendices, and various modifications are possible within the scope of the claims.

[0258] (Appendix A1) a first acquisition means for acquiring inference data and information relating to the type of the inference data; a second acquisition means for identifying similar data that is similar to the inference data by referring to the information related to the type, and acquiring a ratio parameter associated with the similar data; The inference data includes: Applying the regression coefficients for each of a plurality of target models; The ratio parameter associated with the similar data is applied as a ratio parameter that defines a ratio of the plurality of target models. a prediction means for deriving a prediction result by an output means for outputting a prediction result by the prediction means; An information processing device comprising:

[0259] (Appendix A2) When the second acquisition means determines that the similar data does not exist, the output means outputs information indicating that a prediction result cannot be derived for the inference data. 10. The information processing device according to claim 1,

[0260] (Appendix A3) The acquisition process by the second acquisition means includes: A process of calculating a similarity from information about the type of each of a plurality of candidate data and information about the type of the inference data; a process of selecting the similar data from the plurality of candidate data by referring at least to the similarity; Contains An information processing device according to appendix A1 or A2.

[0261] (Appendix A4) The acquisition process by the second acquisition means includes: A process of calculating the difference between feature data included in each of a plurality of candidate data and feature data included in the inference data. Contains The process of selecting similar data includes: The difference is further referenced to select the similar data. 10. The information processing device according to claim 9, wherein the information processing device is a

[0262] (Appendix A5) The output means displays graphs defined by the regression coefficients for at least two of the plurality of target models in a manner that allows them to be distinguished from one another. An information processing device according to any one of appendices A1 to A4.

[0263] (Appendix A6) The regression coefficients and the ratio parameters are a regression coefficient calculation process that calculates a regression coefficient for each of a plurality of target models by referring to the ratio parameters and the similar data; and A ratio parameter calculation process for calculating the ratio parameter by referring to the similar data and the regression coefficient. are the regression coefficients and ratio parameters learned by the learning process including An information processing device according to any one of appendices A1 to A5.

[0264] (Appendix A7) The learning process for the regression coefficients and the ratio parameters is a covariance calculation process for calculating covariance parameters of the prior distribution of the latent variables and a covariance matrix of the posterior distribution of the latent variables by referring to the similar data, the ratio parameters, the regression coefficients, and information on the prior distribution of the latent variables; Further comprising: The ratio parameter calculation process is Calculate the ratio parameter by further referring to the covariance matrix of the posterior distribution of the hidden variables. 10. The information processing device according to claim 9, wherein the information processing device is an information processing device according to claim 8.

[0265] (Appendix A8) an acquisition means for acquiring information on the target data, the number of models of a plurality of target models, and the type of the target data; a regression coefficient calculation means for calculating a regression coefficient for each of the plurality of target models by referring to a ratio parameter that defines a ratio of the plurality of target models and the target data; a ratio parameter calculation means for calculating the ratio parameter by referring to the target data and the regression coefficient; a storage means for storing the regression coefficients calculated by the regression coefficient calculation means and the ratio parameters calculated by the ratio parameter calculation means in association with information about the type; An information processing device comprising:

[0266] (Appendix A9) The ratio parameter calculation means calculates the ratio parameter using an update formula including a normalization term according to the type of the target data. 10. The information processing device according to claim 8,

[0267] (Appendix A10) The obtaining means further obtains information regarding a prior distribution of the latent variables; The information processing device includes: Further comprising a covariance calculation means for calculating a covariance parameter of the prior distribution of the latent variables and a covariance matrix of the posterior distribution of the latent variables by referring to the target data, the ratio parameter, the regression coefficient, and information on the prior distribution of the latent variables; The ratio parameter calculation means calculates the ratio parameter by further referring to a covariance matrix of the posterior distribution of the hidden variables. An information processing device according to appendix A8 or A9.

[0268] [Appendix B] This disclosure includes the techniques described in the following appendices. However, the present invention is not limited to the techniques described in the following appendices, and various modifications are possible within the scope of the claims.

[0269] (Appendix B1) a first acquisition process in which at least one processor acquires inference data and information regarding the type of the inference data; a second acquisition process in which the at least one processor refers to information related to the type to identify similar data that is similar to the inference data, and acquires a ratio parameter associated with the similar data; The inference data includes: Applying the regression coefficients for each of a plurality of target models; The ratio parameter associated with the similar data is applied as a ratio parameter that defines a ratio of the plurality of target models. a prediction process by which the at least one processor derives a prediction result; an output process in which the at least one processor outputs a prediction result obtained by the prediction process; An information processing method comprising:

[0270] (Appendix B2) When the second acquisition process determines that the similar data does not exist, the output process outputs information indicating that a prediction result cannot be derived for the inference data. 1. The information processing method described in Appendix B1.

[0271] (Appendix B3) The acquisition process by the second acquisition process includes: a process in which the at least one processor calculates a similarity from information on the type of each of a plurality of candidate data and information on the type of the inference data; a process in which the at least one processor selects the similar data from the plurality of candidate data by at least referring to the similarity; Contains 1. An information processing method according to Appendix B1 or B2.

[0272] (Appendix B4) The acquisition process by the second acquisition process includes: a process in which the at least one processor calculates a difference between feature data included in each of a plurality of candidate data and feature data included in the data for inference; Contains The process of selecting similar data includes: The difference is further referenced to select the similar data. The information processing method described in Appendix B3.

[0273] (Appendix B5) In the output process, the at least one processor displays graphs defined by the regression coefficients for at least two of the plurality of target models in a manner that allows them to be distinguished from one another. 1. An information processing method according to any one of appendices B1 to B4.

[0274] (Appendix B6) The regression coefficients and the ratio parameters are a regression coefficient calculation process in which the at least one processor calculates a regression coefficient for each of a plurality of target models by referring to the ratio parameter and the similar data; and a ratio parameter calculation process in which the at least one processor calculates the ratio parameter by referring to the similar data and the regression coefficients; are the regression coefficients and ratio parameters learned by the learning process including 1. An information processing method according to any one of Appendices B1 to B5.

[0275] (Appendix B7) The learning process for the regression coefficients and the ratio parameters is a covariance calculation process in which the at least one processor calculates covariance parameters of the prior distribution of the latent variables and a covariance matrix of the posterior distribution of the latent variables by referring to the similar data, the ratio parameters, the regression coefficients, and information on the prior distribution of the latent variables; Further comprising: In the ratio parameter calculation process, the at least one processor: Calculate the ratio parameter by further referring to the covariance matrix of the posterior distribution of the hidden variables. An information processing method as described in Appendix B6.

[0276] (Appendix B8) an acquisition process in which the at least one processor acquires information regarding target data, the number of models of a plurality of target models, and the type of the target data; a regression coefficient calculation process in which the at least one processor calculates a regression coefficient for each of the plurality of target models by referring to a ratio parameter that defines a ratio of the plurality of target models and the target data; a ratio parameter calculation process in which the at least one processor calculates the ratio parameter by referring to the target data and the regression coefficient; a storage process in which the at least one processor stores the regression coefficients calculated by the regression coefficient calculation process and the ratio parameters calculated by the ratio parameter calculation process in association with information about the type; An information processing method comprising:

[0277] (Appendix B9) In the ratio parameter calculation process, the at least one processor calculates the ratio parameter using an update formula including a normalization term according to the type of the target data. The information processing method described in Appendix B8.

[0278] (Appendix B10) In the obtaining process, the at least one processor further obtains information regarding a prior distribution of a hidden variable; The information processing device includes: The at least one processor further includes a covariance calculation process for calculating covariance parameters of the prior distribution of the latent variables and a covariance matrix of the posterior distribution of the latent variables by referring to the target data, the ratio parameters, the regression coefficients, and information on the prior distribution of the latent variables; In the ratio parameter calculation process, the at least one processor further refers to a covariance matrix of the posterior distribution of the hidden variables to calculate the ratio parameter. An information processing method according to Appendix B8 or B9.

[0279] [Appendix C] This disclosure includes the techniques described in the following appendices. However, the present invention is not limited to the techniques described in the following appendices, and various modifications are possible within the scope of the claims.

[0280] (Appendix C1) A program that causes a computer to function as an information processing device, the program comprising: a first acquisition means for acquiring inference data and information relating to the type of the inference data; a second acquisition means for identifying similar data that is similar to the inference data by referring to the information related to the type, and acquiring a ratio parameter associated with the similar data; The inference data includes: Applying the regression coefficients for each of a plurality of target models; The ratio parameter associated with the similar data is applied as a ratio parameter that defines a ratio of the plurality of target models. a prediction means for deriving a prediction result by an output means for outputting a prediction result by the prediction means; An information processing program that functions as a

[0281] (Appendix C2) When the second acquisition means determines that the similar data does not exist, the output means outputs information indicating that a prediction result cannot be derived for the inference data. An information processing program as described in Appendix C1.

[0282] (Appendix C3) The acquisition process by the second acquisition means includes: A process of calculating a similarity from information about the type of each of a plurality of candidate data and information about the type of the inference data; a process of selecting the similar data from the plurality of candidate data by referring at least to the similarity; Contains An information processing program according to appendix C1 or C2.

[0283] (Appendix C4) The acquisition process by the second acquisition means includes: A process of calculating the difference between feature data included in each of a plurality of candidate data and feature data included in the inference data. Contains The process of selecting similar data includes: The difference is further referenced to select the similar data. An information processing program as described in Appendix C3.

[0284] (Appendix C5) The output means displays graphs defined by the regression coefficients for at least two of the plurality of target models in a manner that allows them to be distinguished from one another. An information processing program according to any one of appendices C1 to C4.

[0285] (Appendix C6) The regression coefficients and the ratio parameters are a regression coefficient calculation process that calculates a regression coefficient for each of a plurality of target models by referring to the ratio parameters and the similar data; and A ratio parameter calculation process for calculating the ratio parameter by referring to the similar data and the regression coefficient. are the regression coefficients and ratio parameters learned by the learning process including An information processing program according to any one of appendices C1 to C5.

[0286] (Appendix C7) The learning process for the regression coefficients and the ratio parameters is a covariance calculation process for calculating covariance parameters of the prior distribution of the latent variables and a covariance matrix of the posterior distribution of the latent variables by referring to the similar data, the ratio parameters, the regression coefficients, and information on the prior distribution of the latent variables; Further comprising: The ratio parameter calculation process is Calculate the ratio parameter by further referring to the covariance matrix of the posterior distribution of the hidden variables. An information processing program as described in Appendix C6.

[0287] (Appendix C8) The computer an acquisition means for acquiring information on the target data, the number of models of a plurality of target models, and the type of the target data; a regression coefficient calculation means for calculating a regression coefficient for each of the plurality of target models by referring to a ratio parameter that defines a ratio of the plurality of target models and the target data; a ratio parameter calculation means for calculating the ratio parameter by referring to the target data and the regression coefficient; a storage process of storing the regression coefficients calculated by the regression coefficient calculation means and the ratio parameters calculated by the ratio parameter calculation means in association with information about the type; An information processing program that functions as a

[0288] (Appendix C9) The ratio parameter calculation means calculates the ratio parameter using an update formula including a normalization term according to the type of the target data. An information processing program as described in Appendix C8.

[0289] (Appendix C10) The obtaining means further obtains information regarding a prior distribution of the latent variables; The computer and further functioning as a covariance calculation means for calculating a covariance parameter of the prior distribution of the latent variables and a covariance matrix of the posterior distribution of the latent variables by referring to the target data, the ratio parameter, the regression coefficient, and information on the prior distribution of the latent variables; The ratio parameter calculation means calculates the ratio parameter by further referring to a covariance matrix of the posterior distribution of the hidden variables. An information processing program according to Appendix C8 or C9.

[0290] [Appendix D] This disclosure includes the techniques described in the following appendices. However, the present invention is not limited to the techniques described in the following appendices, and various modifications are possible within the scope of the claims.

[0291] (Appendix D1) at least one processor, a first acquisition process for acquiring inference data and information regarding the type of the inference data; a second acquisition process of identifying similar data that is data similar to the inference data by referring to the information related to the type, and acquiring a ratio parameter associated with the similar data; The inference data includes: Applying the regression coefficients for each of a plurality of target models; The ratio parameter associated with the similar data is applied as a ratio parameter that defines a ratio of the plurality of target models. a prediction process for deriving a prediction result by an output process for outputting a prediction result obtained by the prediction process; An information processing device that executes the above.

[0292] The information processing device may further include a memory, and the memory may store a program for causing the at least one processor to execute each of the processes.

[0293] (Appendix D2) When the second acquisition process determines that the similar data does not exist, the output process outputs information indicating that a prediction result cannot be derived for the inference data. 10. The information processing device according to claim 9, wherein the information processing device is a device for processing information.

[0294] (Appendix D3) The acquisition process by the second acquisition process includes: A process of calculating a similarity from information about the type of each of a plurality of candidate data and information about the type of the inference data; a process of selecting the similar data from the plurality of candidate data by referring at least to the similarity; Contains An information processing device according to appendix D1 or D2.

[0295] (Appendix D4) The acquisition process by the second acquisition process includes: A process of calculating the difference between feature data included in each of a plurality of candidate data and feature data included in the inference data. Contains The process of selecting similar data includes: The difference is further referenced to select the similar data. 10. The information processing device according to claim 9, wherein the information processing device is an information processing device according to claim 1, wherein

[0296] (Appendix D5) In the output process, the at least one processor displays graphs defined by the regression coefficients for at least two of the plurality of target models in a manner that allows them to be distinguished from one another. An information processing device according to any one of appendices D1 to D4.

[0297] (Appendix D6) The regression coefficients and the ratio parameters are a regression coefficient calculation process that calculates a regression coefficient for each of a plurality of target models by referring to the ratio parameters and the similar data; and A ratio parameter calculation process for calculating the ratio parameter by referring to the similar data and the regression coefficient. are the regression coefficients and ratio parameters learned by the learning process An information processing device according to any one of appendices D1 to D5.

[0298] (Appendix D7) The learning process for the regression coefficients and the ratio parameters is the at least one processor: a covariance calculation process for calculating covariance parameters of the prior distribution of the latent variables and a covariance matrix of the posterior distribution of the latent variables by referring to the similar data, the ratio parameters, the regression coefficients, and information on the prior distribution of the latent variables; Further execute In the ratio parameter calculation process, the at least one processor: Calculate the ratio parameter by further referring to the covariance matrix of the posterior distribution of the hidden variables. 10. The information processing device according to claim 9, wherein the information processing device is an information processing device according to claim 8.

[0299] (Appendix D8) at least one processor, an acquisition process for acquiring information regarding the target data, the number of models of a plurality of target models, and the type of the target data; a regression coefficient calculation process for calculating a regression coefficient for each of the plurality of target models by referring to a ratio parameter that defines a ratio of the plurality of target models and the target data; a ratio parameter calculation process for calculating the ratio parameter by referring to the target data and the regression coefficient; a storage process of storing the regression coefficients calculated by the regression coefficient calculation process and the ratio parameters calculated by the ratio parameter calculation process in association with information about the type; An information processing device that executes the above.

[0300] (Appendix D9) In the ratio parameter calculation process, the at least one processor calculates the ratio parameter using an update formula including a normalization term according to the type of the target data. 10. The information processing device according to claim 8,

[0301] (Appendix D10) In the obtaining process, the at least one processor further obtains information regarding a prior distribution of a hidden variable; the at least one processor: Further performing a covariance calculation process to calculate a covariance parameter of the prior distribution of the latent variable and a covariance matrix of the posterior distribution of the latent variable by referring to the target data, the ratio parameter, the regression coefficient, and information on the prior distribution of the latent variable; In the ratio parameter calculation process, the at least one processor further refers to a covariance matrix of the posterior distribution of the hidden variables to calculate the ratio parameter. An information processing device according to appendix D8 or D9.

[0302] [Appendix E] This disclosure includes the techniques described in the following appendices. However, the present invention is not limited to the techniques described in the following appendices, and various modifications are possible within the scope of the claims.

[0303] (Appendix E1) A program that causes a computer to function as an information processing device, the program comprising: a first acquisition process for acquiring inference data and information regarding the type of the inference data; a second acquisition process of identifying similar data that is data similar to the inference data by referring to the information related to the type, and acquiring a ratio parameter associated with the similar data; The inference data includes: Applying the regression coefficients for each of a plurality of target models; The ratio parameter associated with the similar data is applied as a ratio parameter that defines a ratio of the plurality of target models. a prediction process for deriving a prediction result by an output process for outputting a prediction result obtained by the prediction process; A non-transitory recording medium on which an information processing program for executing the above is recorded. [Explanation of symbols]

[0304] 1, 1A, 2, 2A, 2B Information processing equipment 11, 21, 22 Acquisition section 12 Regression coefficient calculation section 13 Covariance calculation part 14 Ratio parameter calculation section 15, 24 Output section 16 Initial value determination section 17 Convergence judgment section 18, 23 Prediction Section TD target data (candidate data) RP ratio parameter RC regression coefficient CI Type Information DI distribution information LR learning results PD inference data PR prediction results

Claims

1. a first acquisition means for acquiring inference data and information relating to the type of the inference data; a second acquisition means for identifying similar data that is similar to the inference data by referring to the information about the type, and acquiring a ratio parameter associated with the similar data; The inference data includes: Applying the regression coefficients for each of a plurality of target models; The ratio parameter associated with the similar data is applied as a ratio parameter that defines a ratio of the plurality of target models. a prediction means for deriving a prediction result by an output means for outputting a prediction result by the prediction means; An information processing device comprising:

2. When the second acquisition means determines that the similar data does not exist, the output means outputs information indicating that a prediction result cannot be derived for the inference data. The information processing device according to claim 1 .

3. The acquisition process by the second acquisition means includes: A process of calculating a similarity from information about the type of each of a plurality of candidate data and information about the type of the inference data; a process of selecting the similar data from the plurality of candidate data by referring at least to the similarity; Contains The information processing device according to claim 2 .

4. The acquisition process by the second acquisition means includes: A process of calculating the difference between feature data included in each of a plurality of candidate data and feature data included in the inference data. Contains The process of selecting similar data includes: The difference is further referenced to select the similar data. The information processing device according to claim 3 .

5. The output means displays graphs defined by the regression coefficients for at least two of the plurality of target models in a manner that allows them to be distinguished from one another. The information processing device according to claim 1 .

6. The regression coefficients and the ratio parameters are a regression coefficient calculation process that calculates a regression coefficient for each of a plurality of target models by referring to the ratio parameters and the similar data; and A ratio parameter calculation process for calculating the ratio parameter by referring to the similar data and the regression coefficient. are the regression coefficients and ratio parameters learned by the learning process including The information processing device according to claim 1 .

7. The learning process for the regression coefficients and the ratio parameters is a covariance calculation process for calculating covariance parameters of the prior distribution of the latent variables and a covariance matrix of the posterior distribution of the latent variables by referring to the similar data, the ratio parameters, the regression coefficients, and information on the prior distribution of the latent variables; Further comprising: The ratio parameter calculation process is Calculate the ratio parameter by further referring to the covariance matrix of the posterior distribution of the hidden variables. The information processing device according to claim 6 .

8. an acquisition means for acquiring information on the target data, the number of models of a plurality of target models, and the type of the target data; a regression coefficient calculation means for calculating a regression coefficient for each of the plurality of target models by referring to a ratio parameter that defines a ratio of the plurality of target models and the target data; a ratio parameter calculation means for calculating the ratio parameter by referring to the target data and the regression coefficient; a storage means for storing the regression coefficients calculated by the regression coefficient calculation means and the ratio parameters calculated by the ratio parameter calculation means in association with information about the type; An information processing device comprising:

9. acquiring inference data and information regarding the type of the inference data; Identifying similar data that is similar to the inference data by referring to the information about the type, and acquiring a ratio parameter associated with the similar data; The inference data includes: Applying the regression coefficients for each of a plurality of target models; The ratio parameter associated with the similar data is applied as a ratio parameter that defines a ratio of the plurality of target models. and deriving a prediction result by outputting the prediction result; An information processing method comprising:

10. 2. A program for causing a computer to function as the information processing device according to claim 1, the program causing a computer to function as the first acquisition means, the second acquisition means, the prediction means, and the output means.

Images