Learning device

Abstract

This learning device comprises: an acquisition unit that acquires a time series data set including time series data pertaining to a prescribed observation item; an extraction unit that uses a model having undergone machine training to extract a feature amount from the time series data set acquired by the acquisition unit; a learning unit that performs machine learning to update a weight of the model using the feature amount extracted by the extraction unit; a detection unit that detects a change in an input dimension, which is a change in an observation item included in the time series data set; and an adjustment unit that adjusts a weight dimension of the model in accordance with the result of detection by the detection unit. According to such a configuration, it is possible to accurately handle even a case where an input dimension changes, and it is possible to suitably support, for example, decision-making by a user even when the input dimension changes.

Description

Learning Device

[0001] The present invention relates to a learning device, a learning method, and a recording medium.

[0002] There is a known technology that enables the system status to be grasped, such as detecting anomalies, by making time-series data searchable.

[0003] For example, Patent Document 1 describes a time series data processing device that converts a time series data set, which is a collection of multiple time series data, into a feature vector that indicates characteristics of the time series data set, thereby enabling search. Specifically, Patent Document 1 describes a time series data processing device that includes a data conversion unit, a storage unit, and a search unit. For example, the data conversion unit converts partial time series data sets, which are obtained by dividing a time series data set, which is a collection of multiple time series data, into feature vectors that indicate characteristics of the partial time series data sets. The storage unit also stores multiple first partial time series data sets and multiple first feature vectors converted from the multiple first partial time series data sets by the data conversion unit, in association with each other. The search unit then selects at least one first feature vector from the multiple first feature vectors stored in the storage unit that is similar to a second feature vector converted from an input second partial time series data set by the data conversion unit, and outputs the first partial time series data set corresponding to the selected first feature vector.

[0004] International Publication No. 2020 / 049666

[0005] The technology described in Patent Document 1 is based on the premise that the input dimension does not change between learning and searching. However, the input dimension may change between learning and searching due to the addition of new observation items or the deletion of existing observation items. As a result, a new model must be prepared, and re-learning is required. Thus, there has been a problem in that it is difficult to accommodate changes in the input dimension when making time-series data searchable.

[0006] Therefore, one of the objects of the present invention is to provide a learning device, a learning method, and a recording medium that can solve the above-mentioned problems.

[0007] In order to achieve this object, a learning device that is one form of the present disclosure has an acquisition unit that acquires a time series dataset including time series data for predetermined observation items; an extraction unit that uses a machine-learned model to extract features from the time series dataset acquired by the acquisition unit; a learning unit that performs machine learning to update the weights of the model using the features extracted by the extraction unit; a detection unit that detects changes in input dimensions, which are changes in observation items included in the time series dataset; and an adjustment unit that adjusts the weight dimensions of the model depending on the results of detection by the detection unit.

[0008] Furthermore, a learning method according to another aspect of the present disclosure includes the steps of: acquiring a time series dataset including time series data on predetermined observation items; extracting features from the acquired time series dataset using a machine-learned model; performing machine learning to update weights of the model using the extracted features; detecting a change in an input dimension, which is a change in an observation item included in the time series dataset; and adjusting a weight dimension of the model according to a result of the detection.

[0009] Furthermore, a recording medium as another aspect of the present disclosure is a computer-readable recording medium having recorded thereon a program for causing an information processing device to perform the following processes: acquire a time series dataset including time series data for predetermined observation items; extract features from the acquired time series dataset using a machine-learned model; perform machine learning to update the weights of the model using the extracted features; detect changes in input dimensions, which are changes in observation items included in the time series dataset; and adjust the weight dimensions of the model according to the detection results.

[0010] According to the above-mentioned configurations, it is possible to cope with changes in input dimensions.

[0011] FIG. 1 is a diagram illustrating an example of a problem when making time-series data searchable. FIG. 2 is a block diagram illustrating an example of a configuration of a learning device. FIG. 3 is a diagram illustrating an example of a process when increasing input dimensions. FIG. 4 is a diagram illustrating an example of a process when decreasing input dimensions. FIG. 5 is a flowchart illustrating an example of an operation of a learning device. FIG. 6 is a block diagram illustrating another example of a configuration of a learning device. FIG. 7 is a diagram illustrating an example of an effect of the present disclosure. FIG. 8 is a diagram illustrating an example of an application of a learning device. FIG. 9 is a diagram illustrating an example of a hardware configuration of a learning device according to a second embodiment of the present disclosure. FIG. 10 is a block diagram illustrating an example of a configuration of a learning device. FIG. 11 is a flowchart illustrating an example of an operation of a learning device.

[0012] [First Embodiment] A first embodiment of the present invention will be described with reference to FIGS. 1 to 8. FIG. 1 is a diagram illustrating an example of a problem when making time-series data searchable. FIG. 2 is a block diagram showing an example of the configuration of a learning device 100. FIGS. 3 and 4 are diagrams for explaining an example of processing when reducing input dimensions. FIG. 5 is a flowchart showing an example of the operation of the learning device 100. FIG. 6 is a block diagram showing another example of the configuration of a learning device. FIG. 7 is a diagram for explaining an example of the effect of the present disclosure. FIG. 8 is a diagram showing an example of application of the learning device 100. Note that in the present disclosure, the drawings may be associated with one or more embodiments.

[0013] In a first embodiment of the present disclosure, a learning device 100 is described, which is an information processing device that learns or updates a machine-learned model based on a time-series dataset including one or more pieces of time-series data, which are time-series observation data. For example, the input dimensions may change from the time of model learning due to the addition of new observation items or the deletion of existing observation items. In such a case, as shown in FIG. 1 , if a new model corresponding to the changed number of input dimensions is prepared, each weight included in the model will be randomly initialized, and re-learning will be necessary from the beginning.

[0014] Therefore, when the learning device 100 described in this disclosure detects a change in the input dimension, it changes the weight dimension of the first layer (input gate) in response to the detected change. In other words, the learning device 100 expands or contracts the weights in response to the change in the input dimension, and continues to use the common part. This allows the learning device 100 to respond to the change in the input dimension without preparing a new model.

[0015] In the present disclosure, a change in input dimension refers to a change in the number of time series data included in a time series dataset due to, for example, adding a new observation item or deleting an existing observation item. In other words, a change in input dimension refers to a change in the number of observation items. For example, the number of time series data included in a time series dataset changes in response to a change in the number of sensors that acquire time series data or a change in the number of data input as time series data. In other words, the input dimension changes.

[0016] In addition, in the present disclosure, it is assumed that the observation content of common observation items remains unchanged before and after a change in the input dimension. In other words, it is assumed that observation items other than the added or deleted observation item continue to acquire time-series data unchanged from before the addition or deletion of the observation item. By changing the weighting dimension in accordance with the change in the input dimension, the learning device 100 can take into account the influence of the change in the observation item while maintaining the relationship between common inputs before and after the change.

[0017] Furthermore, the learning device 100 described in this disclosure can learn a model by using a supervised distance learning method such as that described in Patent Document 1. For example, the learning device 100 randomly extracts anchor samples from a feature group consisting of multiple feature amounts extracted from a time-series dataset, and extracts positive samples assigned to the same class as the anchor samples and negative samples assigned to a different class from the anchor samples. The learning device 100 then learns and updates the model by updating weight parameters so as to reduce the distance between the anchor samples and positive samples and increase the distance between the anchor samples and negative samples in the feature space.

[0018] Fig. 2 shows an example configuration of learning device 100. Referring to Fig. 2, learning device 100 has, as main components, for example, an operation input unit 110, a screen display unit 120, a communication I / F unit 130, a storage unit 140, and an arithmetic processing unit 150.

[0019] 2 illustrates a case where the functions of learning device 100 are realized using a single information processing device. However, at least some of the functions of learning device 100 may be realized using multiple information processing devices, for example, on the cloud. Furthermore, learning device 100 may not include some of the components illustrated above, such as not having operation input unit 110 or screen display unit 120, or may have components other than those illustrated above.

[0020] Operation input unit 110 is made up of operation input devices such as a keyboard, a mouse, etc. Operation input unit 110 detects operations of the operator operating learning device 100 and outputs the operations to calculation processing unit 150.

[0021] The screen display unit 120 is composed of a screen display device such as a liquid crystal display, an organic electroluminescence (EL) display, etc. The screen display unit 120 can display various information stored in the storage unit 140 on the screen in response to instructions from the arithmetic processing unit 150.

[0022] The communication I / F unit 130 is composed of a data communication circuit, etc. The communication I / F unit 130 performs data communication with various sensors and other external devices connected via communication lines.

[0023] The storage unit 140 is a storage device such as a hard disk or memory. The storage unit 140 stores processing information and programs 144 required for various processes in the arithmetic processing unit 150. The programs 144 are read into the arithmetic processing unit 150 and executed to realize various processing units. The programs 144 are read in advance from an external device or recording medium via a data input / output function such as the communication I / F unit 130, and are stored in the storage unit 140. Main information stored in the storage unit 140 includes, for example, model information 141, feature amount information 142, and time-series data information 143.

[0024] The model information 141 includes information about a model that extracts and outputs features that preserve local distance relationships for input time series segments obtained by dividing a time series dataset. For example, the model information 141 may include weight parameters included in the trained model described above. For example, the model included in the model information 141 is trained in advance using training time series segments inside or outside the training device 100 and stored in the storage unit 140. The model information 141 is updated through training by the model training unit 154, which will be described later. The number of dimensions of the weight parameters included in the model information 141 is adjusted by the dimension adjustment unit 156, which will be described later.

[0025] The feature information 142 includes information corresponding to features extracted from the time-series segments. For example, the feature information 142 includes binary codes obtained by converting features using any method. In the feature information 142, the binary codes may be associated with classes to be classified. For example, the binary codes included in the feature information 142 are obtained by converting features extracted by the feature extraction unit 152 using the binary code conversion unit 153, and are stored in the storage unit 140.

[0026] The information included in the feature amount information 142 can be used when performing a search process, which will be described later. For example, the results of the search process may be used when determining whether or not an abnormality exists or to what class the abnormality belongs. The results of the search process may also be used for purposes other than those exemplified above.

[0027] The time-series data information 143 includes a time-series data set consisting of one or more pieces of time-series data. Here, the time-series data may be data in which numerical data, such as observation data measured by a sensor at a predetermined interval, is arranged in order of measurement time. The time-series data may be time-series information input using the operation input unit 110, such as information indicating meal times and contents, body temperature, medication contents, etc. The time-series data information 143 is updated in response to the time-series data acquisition unit 151, which will be described later, acquiring time-series data from sensors and other devices.

[0028] The arithmetic processing unit 150 has an arithmetic device such as a CPU (Central Processing Unit) and its peripheral circuits. The arithmetic processing unit 150 reads and executes a program 144 from the storage unit 140, thereby causing the hardware and the program 144 to work together to realize various processing units. Major processing units realized by the arithmetic processing unit 150 include, for example, a time-series data acquisition unit 151, a feature extraction unit 152, a binary code conversion unit 153, a model learning unit 154, an input dimension change detection unit 155, and a dimension adjustment unit 156.

[0029] In addition, instead of the above-mentioned CPU, the arithmetic processing unit 150 may have a GPU (Graphics Processing Unit), a DSP (Digital Signal Processor), an MPU (Micro Processing Unit), an FPU (Floating point number Processing Unit), a PPU (Physics Processing Unit), a TPU (Tensor Processing Unit), a quantum processor, a microcontroller, or a combination of these.

[0030] The time-series data acquisition unit 151 acquires a time-series data set by acquiring time-series data from various sensors and other devices. The time-series data acquisition unit 151 may accept input of information using the operation input unit 110 or the like as at least a part of the time-series data included in the time-series data set. Furthermore, the time-series data acquisition unit 151 stores the acquired time-series data set in the storage unit 140 as time-series data information 143.

[0031] The feature extraction unit 152 extracts features from the time series segments by inputting the time series segments to a model stored as the model information 141. For example, the feature extraction unit 152 divides the time series data set acquired by the time series data acquisition unit 151 into multiple time series segments using a time window of a certain period. Then, the feature extraction unit 152 inputs each of the divided time series segments to a trained model to extract features. Note that the feature extraction unit 152 may divide the time series data into the time series segments using any method. For example, the size of the time window may be set arbitrarily. Furthermore, the feature extraction unit 152 may divide the time series data into multiple time series segments so that they overlap for an arbitrary period, or may divide the time series data into multiple time series segments so that the time series segments do not overlap.

[0032] The binary code conversion unit 153 converts the features extracted by the feature extraction unit 152 into binary code, which is information corresponding to the features. In the present disclosure, the method of conversion into binary code is not particularly limited. The binary code conversion unit 153 may convert the features extracted by the feature extraction unit 152 into binary code using any method. Furthermore, the binary code conversion unit 153 can store the converted binary code in the storage unit 140 as feature information 142.

[0033] The feature extraction unit 152 or the binary code conversion unit 153 may assign a class to be classified to the feature or binary code using any method. For example, the feature extraction unit 152 or the binary code conversion unit 153 may assign a class to the feature, etc., determined based on the Euclidean distance between the time-series segment from which the feature was extracted and the time-series segment to be compared. Information about the class may be acquired based on an operation by the operator on the operation input unit 110, etc.

[0034] The model learning unit 154 performs learning based on the feature amounts extracted by the feature amount extraction unit 152 , thereby updating the model stored as the model information 141 .

[0035] For example, the model training unit 154 trains the model using a supervised metric learning method such as that described in Patent Literature 1. The model training unit 154 can train the model so as to minimize the metric learning loss. As an example, the model training unit 154 randomly extracts anchor samples from a feature group including multiple feature amounts extracted by the feature extraction unit 152, and extracts positive samples assigned to the same class as the anchor sample and negative samples assigned to a different class from the anchor sample. In this case, the model training unit 154 extracts anchor samples, positive samples, and negative samples from feature amounts included in the feature group that have the same data characteristics as the original samples. The model training unit 154 then updates the model by comparing the extracted samples. For example, the model training unit 154 can update the model by updating weight parameters so as to reduce the distance between the anchor sample and the positive sample and increase the distance between the anchor sample and the negative sample in the feature space.

[0036] The class corresponding to each feature may be assigned by the feature extraction unit 152 or the binary code conversion unit 153, or may be assigned by the model learning unit 154. For example, the model learning unit 154 may assign a class to each feature using a method similar to that used by the feature extraction unit 152 or the binary code conversion unit 153.

[0037] The loss function used by the model learning unit 154 during learning may be any function that preserves local similarity between data in the input space. For example, the loss function may be triplet loss, pairwise loss, contrastive loss, or the like. The model learned by the model learning unit 154 may be any model that can handle time-series data. For example, the model may be one of a 1D-CNN (1-Dimensional-Convolutional Neural Network), a GRU (Gated Recurrent Unit), a LSTM (Long Short-Term Memory), a Transformer, or the like.

[0038] The input dimension change detection unit 155 detects a change in the input dimension. The input dimension change detection unit 155 may use any means to detect a change in the input dimension in response to addition or deletion of an observation item, etc.

[0039] For example, the input dimension change detection unit 155 can detect a change in the input dimension in response to receiving information indicating a change in an observation item, such as an increase or decrease in the number of sensors, from management software that manages the sensors that perform observations. The input dimension change detection unit 155 can also detect a change in the input dimension in response to monitoring the number of time-series data acquired by the time-series data acquisition unit 151. The input dimension change detection unit 155 may detect a change in the input dimension using a method other than those exemplified above, such as detecting a change in the input dimension in response to acquiring information about a new observation item via the operation input unit 110, etc.

[0040] The dimension adjustment unit 156 adjusts the weight dimension of the first layer of the model to be trained, depending on the result of detection by the input dimension change detection unit 155. In other words, the dimension adjustment unit 156 increases or decreases the number of weight parameters of the first layer included in the model information 141, depending on the result of detection by the input dimension change detection unit 155. For example, the dimension adjustment unit 156 can expand the weight dimension in response to detection of an increase in the input dimension, or reduce the weight dimension in response to detection of a decrease in the input dimension.

[0041] For example, referring to FIG. 3, in response to detection of an increase in input dimensions by the input dimension change detection unit 155, such as an increase in the number of sensors, the dimension adjustment unit 156 expands the weight dimension of the first layer. For example, the dimension adjustment unit 156 can add a weight parameter to each node in the first layer according to the number of changes in the input dimensions, such as an increase in the number of sensors. Here, the weight parameter W to be added is new The value of is, for example, a randomly initialized value. As a result, the output of each input gate is affected by the increase in input dimension, such as the influence of newly added sensors, and is calculated as Wx + W new x new In the illustrated formula, W corresponds to a weight parameter, and x corresponds to time-series data acquired from a sensor. new corresponds to an additional weight parameter that is randomly initialized, and x new This corresponds to time series data acquired by sensors with increased weight parameters. new The value of may be arbitrarily determined based on domain knowledge or the like.

[0042] 4, in response to the detection of a decrease in input dimension by the input dimension change detection unit 155, such as a decrease in the number of sensors, the dimension adjustment unit 156 reduces the weight dimension of the first layer. For example, the dimension adjustment unit 156 reduces the number of weight parameters at the node corresponding to the input, such as the decreased sensor. As a result, the output of each input gate is affected by the decrease in input dimension, such as the effect of the decreased sensor, and is reduced by W. Ω x Ω =Wx-W (s-Ω) x (s-Ω)The dimension adjustment unit 156 may identify the corresponding nodes by any means, such as by storing each sensor in association with the corresponding node. The dimension adjustment unit 156 may also adjust the weight parameters for each node in the first layer. In the illustrated formula, s corresponds to the set of sensors before reduction, and Ω corresponds to the set of remaining sensors.

[0043] For example, as exemplified above, the dimension adjustment unit 156 adjusts the weight dimension of the first layer of the model to be trained, in accordance with the result of detection by the input dimension change detection unit 155. Note that the dimension adjustment unit 156 may adjust the weight dimension in the first layer using a method other than the above example, in accordance with the result of detection by the input dimension change detection unit 155. For example, the dimension adjustment unit 156 may adjust the weight dimension using a method according to the type of model to be trained, such as LSTM.

[0044] The above is an example configuration of the learning device 100. With the above configuration, the model learning unit 154 can learn a model of the weight dimension adjusted in accordance with the detection result by the input dimension change detection unit 155. Next, with reference to FIG. 5 , an example of the operation of the learning device 100 when the input dimension changes will be described.

[0045] 5 shows an example of the operation of the learning device 100. Referring to FIG. 5, the input dimension change detection unit 155 detects a change in the input dimension (step S101). The input dimension change detection unit 155 may use any means to detect a change in the input dimension in response to addition or deletion of an observation item, etc.

[0046] When the input dimension change detection unit 155 detects an increase in the input dimension, such as an increase in the number of sensors (step S102, increase), the dimension adjustment unit 156 expands the weight dimension of the first layer. For example, the dimension adjustment unit 156 can add weight parameters to each node in the first layer according to the change in the input dimension, such as an increase in the number of sensors. On the other hand, when the input dimension change detection unit 155 detects a decrease in the input dimension, such as a decrease in the number of sensors (step S102, decrease), the dimension adjustment unit 156 reduces the weight dimension of the first layer. For example, the dimension adjustment unit 156 can reduce the number of weight parameters in the node corresponding to the input, such as the decreased number of sensors.

[0047] The above is an example of the operation of the learning device 100 when the input dimension changes.

[0048] As described above, the learning device 100 includes the input dimension change detection unit 155 and the dimension adjustment unit 156. With this configuration, the dimension adjustment unit 156 can adjust the weighting dimension of the first layer of the model to be learned, in accordance with the detection result by the input dimension change detection unit 155. As a result, the learning device 100 can respond to changes in the input dimension while continuing to use the common part by adjusting the weighting parameters in accordance with changes in the input dimension.

[0049] FIG. 6 shows an example of the effect of the learning device 100 described in the present disclosure. For example, FIG. 6 shows a comparison example between a case where a new model is prepared in response to changes in input dimensions and each weight parameter is randomly initialized, and a case where the learning device 100 is used. In FIG. 6 , the precision is a value such as Precision@1, which indicates the extent to which the features are captured in binary coding. A larger precision value indicates a more accurate and closer conversion. Furthermore, the learning progress is a value such as Triplet loss, which indicates the extent to which learning has progressed. A smaller value indicates a more advanced learning. Referring to FIG. 6 , it can be seen that the case where the learning device 100 is used results in a more accurate and closer conversion than the case where a new model is prepared, and that learning has also progressed.

[0050] The configuration of the learning device 100 is not limited to the configuration exemplified with reference to Fig. 2. For example, the learning device 100 may be configured with only a portion of the configuration exemplified in Fig. 2, such as omitting the binary code conversion unit 153. Fig. 7 shows another example configuration of the learning device 100. Referring to Fig. 7, for example, the arithmetic processing unit 150 can implement a search unit 157 in addition to the configuration exemplified in Fig. 2 by reading and executing the program 144 from the storage unit 140.

[0051] The search unit 157 performs search processing based on the feature amount extracted from the time-series segment to be searched.

[0052] For example, when the time-series data acquisition unit 151 acquires a time-series data set to be searched, the feature extraction unit 152 divides the time-series data set to be searched into multiple time-series segments and extracts features. Furthermore, the binary code conversion unit 153 converts the features extracted by the feature extraction unit 152 into binary codes, which are information corresponding to the features. The search unit 157 acquires the binary codes converted as described above. The search unit 157 then searches the feature information 142 for binary codes similar to the acquired binary code. For example, the search unit 157 calculates the distance between the acquired binary code and each binary code included in the feature information 142. The search unit 157 then acquires, as binary codes similar to the acquired binary code, a binary code that satisfies an arbitrary condition, such as a binary code with the smallest calculated distance, from among the binary codes included in the feature information 142.

[0053] Furthermore, the search unit 157 can perform various processes according to the search results. For example, the search unit 157 can infer the class of the time-series segment to be searched by identifying a class associated with the binary code acquired by the search. Furthermore, the search unit 157 may detect an anomaly according to the search results. For example, the search unit 157 calculates an anomaly score according to the distance between the binary code to be searched and the searched binary code. Then, the search unit 157 can detect an anomaly based on the calculated anomaly score. For example, the search unit 157 may detect an anomaly based on a comparison result between the calculated anomaly score and a predetermined threshold. The search unit 157 may perform processes according to search results other than those exemplified above.

[0054] The learning device 100 described in this disclosure can also be used in various situations where the input dimension may change. For example, the learning device 100 may be used in healthcare settings. For example, as shown in FIG. 8 , the learning device 100 can acquire time-series data such as biometric data using one or more sensors 200 possessed by a patient or other user, or other inputs using the operation input unit 110. In such a configuration, the input dimension may change during use, such as when the types of biometric data acquired, such as heart rate and blood pressure, increase during use due to the addition of sensors 200. Furthermore, the types of biometric data acquired may increase or decrease during sleep, work, eating, exercise, etc. The learning device 100 described in this disclosure can easily adapt to changes in the input dimension during use, such as those exemplified above. In other words, the learning device 100 described in this disclosure can easily perform anomaly detection even when the input dimension changes during use, thereby supporting the decision-making of users such as patients. Furthermore, by realizing a function for sharing detection results with others, such as by transmitting the results to an external device, it is possible to provide better support for decision-making, such as by enabling appropriate intervention by family members or doctors in charge. The application scenarios of the learning device 100 are not limited to the above-mentioned examples. For example, the learning device 100 may be applied in any case where input dimensions other than those exemplified above may change midway, such as when the number of performance monitors linked to the number of paths in an optical network fluctuates during operation.

[0055] Second Embodiment Next, a second embodiment of the present disclosure will be described with reference to Fig. 9 to Fig. 11. Fig. 9 is a diagram illustrating an example of the hardware configuration of a learning device 300. Fig. 10 is a block diagram illustrating an example of the configuration of the learning device 300. Fig. 11 is a flowchart illustrating an example of the operation of the learning device 300.

[0056] In a second embodiment of the present disclosure, a learning device 300 that is an information processing device that performs machine learning using a time-series dataset will be described. FIG. 9 shows an example of the hardware configuration of the learning device 300. Referring to FIG. 9 , the learning device 300 has, as an example, the following hardware configuration: CPU (Central Processing Unit) 301 (arithmetic device); ROM (Read Only Memory) 302 (storage device); RAM (Random Access Memory) 303 (storage device); a group of programs 304 loaded into RAM 303; a storage device 305 that stores the group of programs 304; a drive device 306 that reads and writes data from and to a recording medium 310 external to the information processing device; a communication interface 307 that connects to a communication network 311 external to the information processing device; an input / output interface 308 that inputs and outputs data; and a bus 309 that connects the various components.

[0057] 10 . The program group 304 is stored in advance in the storage device 305 or the ROM 302, for example, and is loaded into the RAM 303 or the like by the CPU 301, and executed by the CPU 301. The program group 304 may be supplied to the CPU 301 via the communication network 311, or may be stored in advance in the recording medium 310, and the drive device 306 may read out the program and supply it to the CPU 301.

[0058] 9 shows an example of the hardware configuration of the learning device 300. The hardware configuration of the learning device 300 is not limited to the above. For example, the learning device 300 may be configured with only a part of the above configuration, such as not including the drive device 306. Furthermore, the CPU 301 may be a GPU, as exemplified in the first embodiment.

[0059] The acquisition unit 321 acquires a time-series data set including time-series data for a predetermined observation item. The acquisition unit 321 may be the time-series data acquisition unit 151 described in the first embodiment, or the like.

[0060] The extraction unit 322 uses a machine-learned model to extract features from the time-series data set acquired by the acquisition unit 321. The extraction unit 322 may be the feature extraction unit 152 described in the first embodiment, or the like. The learning unit 323 performs machine learning to update the weights of the model using the features extracted by the extraction unit 322. The learning unit 323 may be the model learning unit 154 described in the first embodiment, or the like.

[0061] The detection unit 324 detects a change in the input dimension, which is a change in the observation item included in the time-series data set. The detection unit 324 may detect a change in the input dimension by acquiring management information or the like from an external device or by monitoring the time-series data set acquired by the acquisition unit 321. The detection unit 324 may be the input dimension change detection unit 155 described in the first embodiment, or the like.

[0062] The adjustment unit 325 adjusts the weight dimension of the model in accordance with the result of detection by the detection unit 324. The adjustment unit 325 may be the dimension adjustment unit 156 described in the first embodiment, or the like.

[0063] The above is an example of the configuration of the learning device 300. Next, an example of the operation of the learning device 300 when the input dimension changes will be described with reference to FIG.

[0064] 11 is a flowchart showing an example of the operation of the learning device 300. Referring to FIG. 11, the detection unit 324 detects a change in the input dimension, which is a change in the observation item included in the time-series data set (step S201). The detection unit 324 may detect the change in the input dimension using any means.

[0065] The adjustment unit 325 adjusts the weighting dimension of the model according to the result of detection by the detection unit 324 (step S202).

[0066] The above is an example of the operation of the learning device 300 when the input dimension changes.

[0067] As described above, learning device 300 includes detection unit 324 and adjustment unit 325. With this configuration, adjustment unit 325 can adjust the weighting dimension of the model in accordance with the detection result by detection unit 324. As a result, learning device 300 can respond to changes in the input dimension while continuing to use the common part by adjusting the weighting dimension in accordance with changes in the input dimension.

[0068] The above-described learning device 300 can be realized by incorporating a predetermined program into an information processing device such as the learning device 300. Specifically, a program according to another aspect of the present invention is a program for causing an information processing device such as the learning device 300 to acquire a time-series dataset including time-series data on predetermined observation items, perform machine learning to extract features from the acquired time-series dataset using a machine-learned model, and update the weights of the model using the extracted features, as well as detect changes in the input dimension, which is a change in the observation items included in the time-series dataset, and adjust the weight dimension of the model according to the detection result.

[0069] Furthermore, a learning method executed by an information processing device such as the above-described learning device 300 is a method in which the information processing device acquires a time-series dataset including time-series data for predetermined observation items, extracts features from the acquired time-series dataset using a machine-learned model, performs machine learning to update the weights of the model using the extracted features, detects changes in the input dimension, which is a change in the observation items included in the time-series dataset, and adjusts the weight dimension of the model according to the detection results.

[0070] Even if the invention is a program having the above-mentioned configuration, or a computer-readable recording medium on which the program is recorded, or a learning method, it can achieve the same functions and effects as the above-mentioned learning device 300, and therefore can achieve the above-mentioned objective of the present disclosure.

[0071] <Supplementary Notes> Part or all of the above-described embodiments can be described as follows: Below, an outline of the learning device and the like according to the present invention will be described. However, the present invention is not limited to the following configuration.

[0072] (Supplementary Note 1) A learning device comprising: an acquisition unit that acquires a time-series dataset including time-series data for predetermined observation items; an extraction unit that uses a machine-learned model to extract features from the time-series dataset acquired by the acquisition unit; a learning unit that performs machine learning to update model weights using the features extracted by the extraction unit; a detection unit that detects a change in an input dimension, which is a change in an observation item included in the time-series dataset; and an adjustment unit that adjusts a weight dimension of the model in accordance with a result of detection by the detection unit. (Supplementary Note 2) The learning device according to Supplementary Note 1, wherein the adjustment unit adjusts a weight dimension in a first layer of the model. (Supplementary Note 3) The learning device according to Supplementary Note 1 or Supplementary Note 2, wherein the adjustment unit expands a weight dimension of the model in accordance with detection by the detection unit of an increase in input dimension. (Supplementary Note 4) The learning device according to any one of Supplements 1 to 3, wherein the adjustment unit reduces a weight dimension of the model in accordance with detection by the detection unit of a decrease in input dimension. (Supplementary Note 5) The learning device according to Supplementary Note 4, wherein the adjustment unit reduces a weight dimension at a node in a first layer of the model that corresponds to a reduced input. (Supplementary Note 6) The learning device according to any one of Supplements 1 to 5, wherein the detection unit detects a change in the number of sensors that acquire time-series data as a change in the input dimension. (Supplementary Note 7) The learning device according to any one of Supplements 1 to 6, comprising: a conversion unit that converts the feature quantity extracted by the extraction unit into a binary code; and a storage unit that stores the binary code converted by the conversion unit. (Supplementary Note 8) The learning device according to Supplementary Note 7, comprising: a search unit that searches for binary codes stored in the storage unit by using binary codes converted from a time-series dataset to be searched.(Supplementary Note 9) A learning method in which an information processing device acquires a time-series dataset including time-series data on predetermined observation items, extracts features from the acquired time-series dataset using a machine-learned model, performs machine learning to update weights of the model using the extracted features, detects changes in input dimensions that are changes in observation items included in the time-series dataset, and adjusts weight dimensions of the model according to the detection results. (Supplementary Note 10) A computer-readable recording medium having recorded thereon a program for implementing processes in an information processing device: acquires a time-series dataset including time-series data on predetermined observation items, extracts features from the acquired time-series dataset using a machine-learned model, performs machine learning to update weights of the model using the extracted features, detects changes in input dimensions that are changes in observation items included in the time-series dataset, and adjusts weight dimensions of the model according to the detection results.

[0073] Note that some or all of the configurations described in Supplementary Notes 2 to 8 that are dependent on the learning device described in Supplementary Note 1 may also be dependent in a similar dependent relationship on the learning method described in Supplementary Note 9 and the recording medium described in Supplementary Note 10. Furthermore, not limited to Supplementary Notes 9 and 10, some or all of the configurations described as Supplements may also be dependent on various hardware, software, various recording means for recording software, or systems within the scope of the above-described embodiments.

[0074] The programs described in the above embodiments and appendices may be stored in a storage device or a computer-readable recording medium, such as a portable medium such as a flexible disk, an optical disk, a magneto-optical disk, or a semiconductor memory.

[0075] Although the present invention has been described above with reference to the above-mentioned embodiments, the present invention is not limited to the above-mentioned embodiments. Various modifications that can be understood by those skilled in the art can be made to the configuration and details of the present invention within the scope of the present invention.

[0076] 100 Learning device 110 Operation input unit 120 Screen display unit 130 Communication I / F unit 140 Storage unit 141 Model information 142 Feature information 143 Time series data information 144 Program 150 Arithmetic processing unit 151 Time series data acquisition unit 152 Feature extraction unit 153 Binary code conversion unit 154 Model learning unit 155 Input dimension change detection unit 156 Dimension adjustment unit 157 Search unit 200 Sensor 300 Learning device 301 CPU 302 ROM 303 RAM 304 Program group 305 Storage device 306 Drive device 307 Communication interface 308 Input / output interface 309 Bus 310 Recording medium 311 Communication network 321 Acquisition unit 322 Extraction unit 323 Learning unit 324 Detection unit 325 Adjustment unit

Claims

1. A learning device comprising: an acquisition unit that acquires a time series dataset including time series data for specified observation items; an extraction unit that uses a machine-learned model to extract features from the time series dataset acquired by the acquisition unit; a learning unit that performs machine learning to update model weights using the features extracted by the extraction unit; a detection unit that detects changes in an input dimension, which is a change in an observation item included in the time series dataset; and an adjustment unit that adjusts the weight dimension of the model depending on a result of detection by the detection unit.

2. A learning device according to claim 1, wherein the adjustment unit adjusts a weight dimension in a first layer of the model.

3. A learning device according to claim 1, wherein the adjustment unit expands a weight dimension of the model in response to detection by the detection unit of an increase in the input dimension.

4. A learning device according to claim 1, wherein the adjustment unit reduces a weight dimension of the model in response to detection by the detection unit of a reduction in the input dimension.

5. A learning device according to claim 4, wherein the adjustment unit reduces a weight dimension in a node corresponding to a reduced input in a first layer of the model.

6. A learning device according to claim 1, wherein the detection unit detects a change in the number of sensors acquiring time-series data as a change in the input dimension.

7. A learning device as claimed in claim 1, comprising: a conversion unit that converts the feature quantity extracted by the extraction unit into a binary code; and a memory unit that stores the binary code converted by the conversion unit.

8. A learning device according to claim 7, further comprising a search unit which searches for binary codes stored in said memory unit using binary codes obtained by converting a time series data set to be searched.

9. A learning method in which an information processing device acquires a time series dataset including time series data for specified observation items, performs machine learning to extract features from the acquired time series dataset using a machine-learned model, and updates the weights of the model using the extracted features, while detecting a change in an input dimension, which is a change in an observation item included in the time series dataset, and adjusts the weight dimension of the model depending on the detection result.

10. A computer-readable recording medium having recorded thereon a program for implementing the following processes in an information processing device: acquiring a time series dataset including time series data for specified observation items; extracting features from the acquired time series dataset using a machine-learned model; performing machine learning to update the weights of the model using the extracted features; detecting changes in input dimensions, which are changes in observation items included in the time series dataset; and adjusting the weight dimensions of the model depending on the detection results.

Images