Information processing device, information processing method, and program

Abstract

This information processing device is provided with: an acquisition means for acquiring monomer information that is information pertaining to one or a plurality of monomers; and a training means for executing a training process pertaining to a prediction model that predicts a polymer material including one or a plurality of constituent units that constitute the one or a plurality of monomers, the training process referring at least to the monomer information acquired by the acquisition means, and a formula formulated on the basis of a physical law or information obtained by using the formula.

Description

Information processing device, information processing method, and program

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

[0002] Techniques using machine learning models to predict the physical properties of compounds composed of multiple molecules are known. For example, Patent Literature 1 discloses a physical property prediction device that predicts the physical properties of polymers using a trained model.

[0003] International Publication No. 2022 / 270530

[0004] Generally, in compounds such as polymers that are composed of many molecules, the molecular chains are long and the interrelationships between the molecules are complex. Therefore, even if the technology of Patent Document 1 is used, the computational cost for constructing a machine learning model is very high.

[0005] The present disclosure has been made in consideration of the above-described problems, and an exemplary purpose thereof is to provide a technology that can suitably perform predictions regarding polymer materials or construct prediction models for making such predictions while suppressing calculation costs.

[0006] An information processing device according to an exemplary aspect of the present disclosure includes an acquisition means for acquiring monomer information, which is information relating to one or more monomers, and a learning means for executing a learning process for a prediction model that makes predictions regarding a polymeric material including one or more structural units that constitute one or more monomers, by referring at least to the monomer information acquired by the acquisition means and an equation formulated based on physical laws or information obtained using the equation.

[0007] An information processing device according to an exemplary aspect of the present disclosure includes an acquisition means for acquiring monomer information, which is information relating to one or more monomers, and a prediction means for generating a prediction result by inputting at least the monomer information acquired by the acquisition means into a prediction model that makes predictions regarding a polymeric material including one or more structural units that constitute one or more monomers, wherein the prediction model is a model constructed by a learning process that at least references the monomer information for learning and an equation formulated based on physical laws, or information obtained using the equation.

[0008] An information processing method according to an exemplary aspect of the present disclosure includes acquiring monomer information, which is information relating to one or more monomers, and performing a learning process for a prediction model that makes predictions regarding a polymeric material including one or more structural units that constitute the one or more monomers, by referring to at least the acquired monomer information and an equation formulated based on physical laws or information obtained using the equation.

[0009] An information processing method according to an exemplary aspect of the present disclosure includes acquiring monomer information, which is information relating to one or more monomers, and generating a prediction result by inputting at least the acquired monomer information into a prediction model that makes predictions regarding a polymeric material including one or more structural units that constitute one or more monomers, wherein the prediction model is a model constructed by a learning process that at least references the monomer information for learning and an equation formulated based on physical laws, or information obtained using the equation.

[0010] A program according to an exemplary aspect of the present disclosure is a program that causes a computer to function as an information processing device, and the program causes the computer to execute an acquisition process that acquires monomer information, which is information about one or more monomers, and further executes a learning process for a prediction model that makes predictions about a polymeric material containing one or more structural units that constitute one or more monomers, by referring at least to the monomer information acquired by the acquisition process and an equation formulated based on physical laws or information obtained using the equation.

[0011] A program according to an exemplary aspect of the present disclosure is a program that causes a computer to function as an information processing device, and the program causes the computer to execute an acquisition process that acquires monomer information, which is information about one or more monomers, and a prediction process that generates a prediction result by inputting at least the monomer information acquired by the acquisition process into a prediction model that makes predictions about a polymeric material including one or more structural units that constitute one or more monomers, and the prediction model is a model constructed by a learning process that at least references the monomer information for learning and an equation formulated based on the laws of physics, or information obtained using the equation.

[0012] According to one exemplary aspect of the present disclosure, an exemplary effect is achieved in that a technology can be provided that can suitably perform predictions regarding polymer materials or construct prediction models for making such predictions while suppressing computational costs.

[0013] FIG. 1 is a block diagram showing the configuration of an information processing device according to an exemplary embodiment. FIG. 2 is a flow diagram showing the flow of an information processing method according to an exemplary embodiment. FIG. 3 is a block diagram showing the configuration of an information processing device according to an exemplary embodiment. FIG. 4 is a flow diagram showing the flow of an information processing method according to an exemplary embodiment. FIG. 5 is a block diagram showing the configuration of an information processing device according to an exemplary embodiment. FIG. 6 is a diagram for explaining processing by an information processing device according to an exemplary embodiment. FIG. 7 is a diagram for explaining processing by an information processing device according to an exemplary embodiment. FIG. 8 is a block diagram showing the configuration of an information processing device according to an exemplary embodiment. FIG. 9 is a block diagram showing the configuration of a computer that functions as an information processing device according to each exemplary embodiment.

[0014] 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 technologies (part or all of the products or methods) 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 technologies 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.

[0015] [First Exemplary Embodiment] A first exemplary embodiment, which is an example of an embodiment of the present invention, will be described in detail with reference to the drawings. This exemplary embodiment is a basic form for each of the exemplary embodiments described below. Note that the scope of application of each technique employed in this exemplary embodiment is not limited to this exemplary embodiment. That is, each technique 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 technique shown in the drawings 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.

[0016] (Configuration of information processing device 1) The configuration of the information processing device 1 according to this exemplary embodiment 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 an acquisition unit 11 and a learning unit 12.

[0017] (Acquisition Unit 11) The acquisition unit 11 acquires monomer information, which is information related to one or more monomers. Here, the monomer information is, for example, information including at least one molecular structure of the monomer and may be expressed in the form of graph information, but this configuration does not limit the present exemplary embodiment. More specifically, the monomer information is information determined according to the molecular structure of the monomer and may include information such as the molecular weight of the monomer, the chemical composition of the monomer, functional groups contained in the monomer, and the electron density distribution in the molecular structure of the monomer, but this does not limit the present exemplary embodiment.

[0018] (Learning unit 12) The learning unit 12 performs a learning process for the prediction model by at least referring to the monomer information acquired by the acquisition unit 11. Here, the prediction model is, for example, a prediction model that makes predictions about a polymer material that includes one or more structural units that constitute one or more monomers, and the structural units are, for example, one or more molecules.

[0019] The learning process performed by the learning unit 12 is performed by, for example, referring to at least the following: - the monomer information acquired by the acquisition unit 11; and - an equation formulated based on the laws of physics, or information obtained using the equation.

[0020] Here, the learning process may include, as an example, a process of updating at least one of the following: one or more parameters included in the prediction model; and one or more parameters included in the "equation formulated based on the laws of physics."

[0021] Furthermore, examples of formulas formulated based on physical laws include: mathematical formulas expressing physical laws based on fundamental interactions such as electromagnetic interactions, gravitational interactions, and nuclear forces, and formulated microscopically based on quantum mechanics; and mathematical formulas expressing physical laws mesoscopically or macroscopically formulated in quantum, classical, or semi-classical mechanics based on the microscopic physical laws. In particular, in this exemplary embodiment, the above-described "formulas formulated based on physical laws" are also referred to as "theoretical formulas," and may include, for example, "formulas (mathematical formulas) formulated based on the physical laws of polymers." However, the above examples do not limit this exemplary embodiment. Furthermore, the term "physical laws" may include "chemical laws," "biochemical laws," and "physicochemical laws." Furthermore, the term "mathematical formulas" in the above description is not limited to equations or inequalities containing equal or inequality signs, but may refer to only one or more terms included in these equations or inequalities.

[0022] Furthermore, in the above description, "information obtained using an equation formulated based on the laws of physics" may include, for example, values ​​substituted for one or more variables included in the equation, or values ​​of one or more variables calculated by the equation. For example, if the "equation formulated based on the laws of physics" can be expressed as z = f(x, y) using variables x, y, and z and a function f, information obtained using the equation may include: (x1, y1), (x2, y2), ..., which are specific sets of values ​​substituted for (x, y); and z1, z2, ..., which are specific values ​​of z obtained when each of (x1, y1), (x2, y2), ... is substituted. In this way, the learning process performed by the learning unit 12 may: directly reference the "equation formulated based on the laws of physics," or may indirectly reference the "equation formulated based on the laws of physics" by referencing the "information obtained using an equation formulated based on the laws of physics."

[0023] (Effects of Information Processing Device 1) As described above, the information processing device 1 is configured to acquire monomer information, which is information relating to one or more monomers, and to execute a learning process for a prediction model that makes predictions about a polymer material containing one or more structural units that constitute one or more monomers, by referring to at least the acquired monomer information and an equation formulated based on physical laws, or information obtained using the equation.

[0024] In this way, the information processing device 1 performs a learning process that at least references an equation formulated based on the laws of physics or information obtained using the equation, and therefore can effectively construct a predictive model that makes predictions about polymer materials while keeping computational costs down.

[0025] (Flow of Information Processing Method S1) Next, the flow of information processing method S1 according to this exemplary embodiment will be described with reference to Fig. 2. Fig. 2 is a flow diagram showing the flow of information processing method S1. As shown in Fig. 2, information processing method S1 includes an acquisition process (acquisition step) S11 and a learning process (learning step) S12. As an example, information processing method S1 is executed by information processing device 1, but this does not limit this exemplary embodiment.

[0026] (Step S11) In step S11, the acquisition unit 11 acquires monomer information, which is information about one or more monomers. The monomer information has been described above, so a description thereof will be omitted here.

[0027] (Step S12) Subsequently, in step S12, the learning unit 12 executes a learning process for the prediction model by at least referring to the monomer information acquired by the acquisition unit 11 in step S11. Here, the prediction model is, for example, a prediction model that makes predictions about a polymer material including one or more structural units that constitute one or more monomers, and the structural units are, for example, molecules.

[0028] The learning process performed by the learning unit 12 in this step is performed by, for example, referring to at least the following: - the monomer information acquired by the acquisition unit 11, and - an equation formulated based on physical laws, or information obtained using the equation. The specific contents of the learning process have been described above, so a detailed description thereof will be omitted here.

[0029] In addition, the above steps S11 and S12 in the information processing method S1 may be repeatedly executed until the prediction model satisfies the desired properties (for example, until one or more parameters included in the prediction model satisfy a predetermined convergence condition).

[0030] (Effects of Information Processing Method S1) As described above, the information processing method S1 is configured to acquire monomer information, which is information related to one or more monomers, and to execute a learning process for a prediction model that makes predictions related to a polymeric material including one or more structural units that constitute one or more monomers, by referring to at least the acquired monomer information and an equation formulated based on physical laws, or information obtained using the equation.

[0031] In this way, in the information processing method S1, a learning process is performed that at least references an equation formulated based on the laws of physics or information obtained using the equation, so that a predictive model that makes predictions about polymer materials can be suitably constructed while suppressing computational costs.

[0032] (Configuration of information processing device 2) Next, the configuration of the information processing device 2 according to this exemplary embodiment 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 and a prediction unit 12.

[0033] (Acquisition Unit 21) The acquisition unit 21 acquires monomer information, which is information about one or more monomers. Here, the monomer information is, for example, information including at least one molecular structure of the monomer and may be expressed in the form of graph information, but this configuration does not limit the present exemplary embodiment.

[0034] (Prediction unit 22) The prediction unit 22 generates a prediction result by inputting at least the monomer information acquired by the acquisition unit 21 into a prediction model that makes predictions regarding a polymer material including one or more structural units that constitute one or more monomers.

[0035] Here, the prediction model is a model constructed by a learning process that references at least: - monomer information for learning; and - an equation formulated based on physical laws, or information obtained using the equation. As an example, the prediction model may be a prediction model trained by the information processing device 1 described above. Note that the "equation formulated based on physical laws" has been described above, and therefore will not be described here.

[0036] (Effects of Information Processing Device 2) As described above, the information processing device 2 acquires monomer information, which is information relating to one or more monomers, and generates a prediction result by inputting at least the acquired monomer information into a prediction model that makes predictions regarding a polymer material containing one or more structural units that constitute one or more monomers, and the prediction model is constructed by a learning process that at least references the monomer information for learning, and an equation formulated based on the laws of physics, or information obtained using the equation.

[0037] In this way, the information processing device 2 generates prediction results using a formula based on the laws of physics or a prediction model constructed by a learning process that at least references information obtained using the formula, so that predictions regarding polymer materials can be made effectively while keeping calculation costs down.

[0038] (Flow of Information Processing Method S2) Next, the flow of information processing method S2 according to this exemplary embodiment will be described with reference to Fig. 4. Fig. 4 is a flow diagram showing the flow of information processing method S2. As shown in Fig. 4, information processing method S2 includes acquisition processing (acquisition step) S21 and prediction processing (learning step) S22. As an example, information processing method S2 is executed by information processing device 2, but this does not limit this exemplary embodiment.

[0039] (Step S21) In step S21, the acquisition unit 21 acquires monomer information, which is information relating to one or more monomers. The monomer information has been described above, so a description thereof will be omitted here.

[0040] (Step S22) Subsequently, in step S22, the prediction unit 22 generates a prediction result by inputting at least the monomer information acquired by the acquisition unit 21 in step S21 into a prediction model that makes predictions regarding a polymer material including one or more structural units that constitute one or more monomers.

[0041] Here, the prediction model is a model constructed by a learning process that references at least: - monomer information for learning; and - an equation formulated based on physical laws, or information obtained using the equation. As an example, the prediction model may be a prediction model trained by the information processing device 2 described above. Note that the "equation formulated based on physical laws" has been described above, and therefore will not be described here.

[0042] (Effects of Information Processing Method S2) As described above, information processing method S2 employs a configuration in which monomer information, which is information relating to one or more monomers, is acquired, and a prediction result is generated by inputting at least the acquired monomer information into a prediction model that makes predictions regarding a polymeric material including one or more structural units that constitute one or more monomers, and the prediction model is constructed by a learning process that at least references the monomer information for learning, and an equation formulated based on the laws of physics, or information obtained using the equation.

[0043] In this way, in the information processing method S2, prediction results are generated using a formula based on the laws of physics or a prediction model constructed by a learning process that at least references information obtained using the formula, so that predictions regarding polymer materials can be made effectively while keeping calculation costs down.

[0044] Second Exemplary Embodiment A second 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 denoted by the same reference numerals, and their description will be omitted as appropriate. The scope of application of each technique employed in this exemplary embodiment is not limited to this exemplary embodiment. That is, each technique 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 hindrance occurs. Furthermore, each technique shown in each drawing referenced to describe this exemplary embodiment can also be employed in other exemplary embodiments included in the present disclosure, to the extent that no particular technical hindrance occurs.

[0045] (Configuration of information processing device 1A) The configuration of information processing device 1A according to this exemplary embodiment will be described with reference to Fig. 5. Fig. 5 is a block diagram showing the configuration of information processing device 1A. As shown in Fig. 5, information processing device 1A includes an acquisition unit 11 and a learning unit 12 that are included in information processing device 1. More specifically, as shown in Fig. 5, information processing device 1A includes a control unit 10A, a storage unit 15A, a communication unit 16A, and an input / output unit 17A.

[0046] (Communication unit 16A) The communication unit 16A communicates with devices external to the information processing device 1A. As an example, the communication unit 16A communicates with one or more terminal devices located outside the information processing device 1A. The communication unit 16A transmits data supplied from the control unit 10A to the terminal devices, and supplies data received from the terminal devices to the control unit 10A.

[0047] (Input / Output Unit 17A) The input / output unit 17A is configured to include at least one of input / output devices such as a keyboard, a mouse, a display, a printer, and a touch panel. Alternatively, the input / output unit 17A may be configured to have input / output devices such as a keyboard, a mouse, a display, a printer, and a touch panel connected to it. In this configuration, the input / output unit 17A accepts various types of information input to the information processing device 1A from the connected input devices. Furthermore, under the control of the control unit 10A, the input / output unit 17A outputs various types of information to the connected output devices. An example of the input / output unit 17A is an interface such as a USB (Universal Serial Bus).

[0048] (Memory Unit 15A) The memory unit 15A stores various data referenced by the control unit 10A and various data generated by the control unit 10A. As an example, the memory unit 15A stores the following: monomer information MI; physical property information PPI; chain information CI; theoretical formula information TFI; and model information MD. Here, the monomer information MI is, as an example, information acquired by the acquisition unit 11 and referenced by the learning unit 12, and includes at least one molecular structure of a monomer. More specifically, the monomer information is information determined according to the molecular structure of a monomer and may include information such as the molecular weight of the monomer, the chemical composition of the monomer, functional groups contained in the monomer, and the electron density distribution in the molecular structure of the monomer, but this does not limit the present exemplary embodiment.

[0049] The physical property information MI is, for example, information acquired by the acquisition unit 11 and referenced by the learning unit 12, and is information related to target physical properties. The physical property information MI includes, for example, information related to physical properties related to the polymer material that is the target of prediction by the prediction model learned by the learning unit 12 (physical properties that define the polymer material). As will be described in detail later, the physical property information MI may include: a correct answer label associated with the monomer information MI, which is a correct answer label related to the physical properties of a polymer material composed of one or more structural units (molecules) that constitute the monomer indicated by the monomer information MI; and information related to the physical properties of the polymer material, which is information related to the physical properties desired by the user.

[0050] The chain information CI is, for example, information acquired by the acquisition unit 11 and referenced by the learning unit 12, and is information regarding a polymer chain composed of one or more structural units (molecules) that constitute a monomer. The chain information CI includes, for example, information regarding at least one of the chain length of the polymer chain composed of the molecules that constitute the monomer, the density of the polymer chain, and the molecular weight distribution of the polymer chain. Here, the "density" may be weight density, electron density distribution, or number density. The chain information CI may also include information regarding the rigidity, polarity, excluded volume, etc. of the polymer chain. Note that the term "polymer chain" in the above description does not limit this exemplary embodiment, and may also be expressed as a "polymer chain," a "molecular chain," etc.

[0051] The theoretical formula information TFI is information referenced by the learning unit 12, and includes, for example, at least: a formula formulated based on the laws of physics, or information obtained using the formula.

[0052] Here, as described in the first exemplary embodiment, the term "formulas formulated based on physical laws" may include, for example, mathematical formulas expressing physical laws based on fundamental interactions such as electromagnetic interactions, gravitational interactions, and nuclear forces, and formulated microscopically based on quantum mechanics, and mathematical formulas expressing physical laws mesoscopically or macroscopically, based on the microscopic physical laws, such as quantum mechanical, classical mechanical, or semi-classical mechanical formulas. In particular, in this exemplary embodiment, the above-described "formulas formulated based on physical laws" are also referred to as "theoretical formulas," and may include, for example, "formulas (mathematical formulas) formulated based on the physical laws of polymers." However, the above examples do not limit this exemplary embodiment. Furthermore, the term "physical laws" may include "chemical laws," "biochemical laws," and "physicochemical laws." Furthermore, the term "mathematical formulas" in the above description is not limited to equations or inequalities containing equal or inequality signs, but may refer to only one or more terms included in these equations or inequalities.

[0053] Furthermore, in the above description, "information obtained using an equation formulated based on the laws of physics" may include, for example, each value substituted into one or more variables included in the equation, or each value of one or more variables calculated by the equation, as described in exemplary embodiment 1. As an example, if the "equation formulated based on the laws of physics" can be expressed as z = f(x, y) using variables x, y, and z and a function f, information obtained using the equation may include: (x1, y1), (x2, y2), ..., which are a set of specific values ​​substituted as (x, y); and z1, z2, ..., which are specific values ​​of z obtained when each of the above (x1, y1), (x2, y2), ... is substituted. As will be described later, the learning process performed by the learning unit 12 may be in the form of: directly referencing the "formula formulated based on the laws of physics," or in the form of indirectly referencing the "formula formulated based on the laws of physics" by referring to "information obtained using the formula formulated based on the laws of physics."

[0054] The model information MD is information that defines a prediction model that is the target of the learning process executed by the learning unit 12. Here, the prediction model is, for example, a model that makes predictions regarding a polymer material that includes one or more structural units that constitute one or more monomers. The model information MD includes, for example, one or more parameters that the prediction model includes.

[0055] Here, a prediction model is information that represents the relationship between explanatory variables and a target variable. A prediction model is, for example, a component that estimates the outcome of an estimation target by calculating a target variable based on the explanatory variables.

[0056] A prediction model may be represented by, for example, a function c that maps an input x to a correct answer y. A prediction model may estimate a numerical value of an estimation target, or may estimate a label of an estimation target. A prediction model may output a variable that describes the probability distribution of a target variable. A prediction model may also be referred to as a "learning model," "analysis model," "AI model," "trained model," "inference model," or "prediction formula."

[0057] Note that explanatory variables are variables used as inputs in a prediction model. Explanatory variables are sometimes referred to as "features" or "features."

[0058] Furthermore, the learning algorithm for generating the predictive model is not particularly limited and may be any existing learning algorithm, such as a random forest, a support vector machine, a naive Bayesian model, a neural network, a piecewise linear model using factorized asymptotic Bayesian inference (FAB), or a neural network.

[0059] (Control Unit 10A) As shown in Fig. 5, the control unit 10A includes an acquisition unit 11 and a learning unit 12. Furthermore, as shown in Fig. 5, the learning unit 12 includes a prediction unit 121 and an update unit 122.

[0060] (Acquisition unit 11) The acquisition unit 11 acquires monomer information, which is information about one or more monomers. Here, the monomer information is, for example, information including at least one molecular structure of the monomer. Details of the monomer information have been described above, so a detailed description thereof will be omitted here.

[0061] (Prediction unit 121) The prediction unit 121 generates a prediction result by inputting monomer information MI to a prediction model defined by model information MD. Here, the prediction model is, for example, a prediction model that makes predictions regarding a polymer material including one or more structural units that constitute one or more monomers. Here, the structural units are, for example, one or more molecules.

[0062] (Update unit 122) The update unit 122 updates the prediction model using the prediction result generated by the prediction unit 121 and the reference result. In other words, the update unit 122 updates one or more parameters of the prediction model using the prediction result generated by the prediction model and the reference result. Here, the reference result refers to, for example, correct answer information (correct answer label) that is compared with the prediction result. As an example, the acquisition unit 11 acquires monomer information MI and a correct answer label associated with the monomer information, the prediction unit 121 generates a prediction result from the monomer information MI using the prediction model, and the update unit 122 updates one or more parameters of the prediction model so as to reduce the difference between the prediction result and the correct answer label.

[0063] As described above, the learning process for the predictive model performed by the learning unit 12 is a learning process that refers to the monomer information MI acquired by the acquisition unit 11, but this learning process also refers to at least an equation formulated based on the laws of physics or information obtained using this equation.

[0064] Such learning processes include at least one of the following: - a case where the prediction model itself is defined using at least an equation formulated based on the laws of physics or information obtained using said equation; - a case where at least an equation formulated based on the laws of physics or information obtained using said equation is used to generate data to be input into the prediction model; and - a case where at least an equation formulated based on the laws of physics or information obtained using said equation is used in an update process to update the prediction model.

[0065] (Example of Processing by Learning Unit 12) Hereinafter, specific examples of processing by the learning unit 12 will be described with reference to different drawings.

[0066] (Processing Example A) FIG. 6 is a diagram showing examples of processes executed by the learning unit 12 according to processing example A, and examples of information referenced or generated in each process.

[0067] 6, in this example, the prediction unit 121 included in the learning unit 12 executes: a conversion process PA1 that converts chain information CI using a theoretical formula; and a prediction process PA2 using a model A that references the result of the conversion process PA1 and the monomer information MI. In other words, the learning process in this example includes a process of inputting the monomer information MI and information obtained by inputting the chain information CI into the formula (theoretical formula) into the prediction model. Here, the model A is an example of a prediction model defined by the above-mentioned model information MD.

[0068] (Conversion Process PA1) As an example, the conversion process PA1 includes at least one of the following: a process of converting at least one of the chain length, density, and molecular weight distribution included in the chain information CI into a descriptor for input to the model A, and a process of converting at least one of the chain length, density, and molecular weight distribution included in the chain information CI into an index or physical quantity for input to the model A.

[0069] Furthermore, the theoretical formula used in the conversion process PA1 may be selected from multiple candidates. For example, the theoretical formula may be selected with reference to at least one of the monomer information MI, the physical property information PPI, and the chain information CI. To achieve this configuration, the learning process of this example may include a selection process for selecting the formula (theoretical formula) from one or more candidates for the formula (theoretical formula) with reference to at least one of the monomer information MI, the physical property information PPI, and the chain information CI. Alternatively, the learning process of this example may include a process for presenting one or more candidates for the formula (theoretical formula) to a user via the input / output unit 17A and selecting the formula (theoretical formula) based on a selection instruction from the user. According to the above configuration, the selection process allows a more suitable theoretical formula to be selected and used in the learning process, thereby enabling a more suitable learning process to be performed. Furthermore, a prediction model using a more suitable theoretical formula can be constructed.

[0070] (Prediction Process PA2) Meanwhile, in prediction process PA2, a prediction result (inference result) is generated by inputting the result of the conversion process by conversion process PA1 and the monomer information MI into model A. The content and type of the prediction result by model A do not limit this exemplary embodiment, but one example can be a prediction result regarding the physical properties of a polymer material that includes one or more structural units that constitute one or more monomers indicated by the monomer information MI.

[0071] 6, the prediction result (inference result) by the prediction process PA2 is provided to the update unit 122. The update unit 122 executes update process P122, using the prediction result by the prediction process PA2 and the reference result, to update at least one of: one or more parameters of the prediction model (prediction model A), and one or more parameters defining the theoretical formula.

[0072] As an example, the update unit 122 executes an update process P122 that updates one or more parameters of the prediction model (prediction model A) or one or more parameters that define the theoretical formula so as to reduce the difference between: the prediction result by the prediction process PA2; and the correct label associated with at least one of the monomer information MI and the chain information CI.

[0073] Here, the correct label is, as an example, a correct label associated with the monomer information MI or the chain information CI, and is a correct label regarding the physical properties of the polymer material composed of the monomer indicated by the monomer information MI or one or more structural units (molecules) that constitute the polymer chain indicated by the chain information CI, but this does not limit this exemplary embodiment.

[0074] Furthermore, the update process P122 may include a process of updating one or more parameters of the prediction model or one or more parameters that define the theoretical formula by referring to a loss function expressed using the theoretical formula indicated by the theoretical formula information TFI. As an example, the above-mentioned conversion process PA1, prediction process PA2, and update process P122 are repeatedly executed until the prediction model satisfies desired properties (for example, until one or more parameters included in the prediction model satisfy predetermined convergence conditions).

[0075] The information processing device 1A according to this example includes an acquisition unit 11 that acquires monomer information and chain information, and a learning means that executes a learning process for a prediction model that makes predictions about a polymer material including one or more structural units that constitute one or more monomers. The learning process can also be expressed as including a process of inputting information obtained by inputting the monomer information acquired by the acquisition unit 11 and the chain information acquired by the acquisition unit 11 into an equation formulated based on physical laws into the prediction model.

[0076] (Processing Example B) FIG. 7 is a diagram showing examples of processes executed by the learning unit 12 according to processing example B, and examples of information referenced or generated in each process.

[0077] As shown in FIG. 7 , in this example, the prediction unit 121 included in the learning unit 12 performs: a conversion process PB1 using a model B1 that references the monomer information MI; and a prediction process PB2 using a model B2 that references the result of the conversion process using the model B1 and at least one of the chain information CI and the theoretical formula information TFI. Here, the models B1 and B2 are examples of prediction models defined by the above-mentioned model information MD. Furthermore, the model B2 can also be expressed as a model incorporating the theoretical formula indicated by the theoretical formula information TFI. In other words, the prediction models used by the prediction unit 121 in this example may be expressed as including: a model B1 (first model) to which the monomer information MI is input; and a model B2 (second model) to which the output of the model B1 (first model) and the chain information CI are input, the second model incorporating the formula (theoretical formula indicated by the theoretical formula information TFI).

[0078] (Conversion Process PB1) The conversion process PB1 includes, as an example, at least one of the following processes: a process of converting the molecular structure of the monomer contained in the monomer information MI into a descriptor for input into the model B2; and a process of converting the molecular structure of the monomer contained in the monomer information MI into an index or physical quantity for input into the model B2.

[0079] (Prediction Process PB2) Meanwhile, in the prediction process PB2, a prediction result (inference result) is generated by inputting the result of the conversion process by the model B1 and at least one of the chain information CI and the theoretical formula information TFI into the model B2. As an example, in the prediction process PB2, a theoretical formula selected from a plurality of theoretical formula candidates with reference to the chain information CI and the result of the conversion process by the model B1 into the model B2 may be configured to generate a prediction result (inference result). To achieve this configuration, the learning process according to this example may include a selection process for selecting the formula (theoretical formula) from one or more candidates for the formula (theoretical formula) with reference to at least one of the monomer information MI, the physical property information PPI, and the chain information CI. Alternatively, the learning process according to this example may include a process for presenting one or more candidates for the formula (theoretical formula) to the user via the input / output unit 17A and selecting the formula (theoretical formula) based on a selection instruction from the user. According to the above configuration, the selection process can select a more suitable theoretical formula to use in the learning process, thereby enabling the execution of a more suitable learning process and the construction of a prediction model using the more suitable theoretical formula.

[0080] Furthermore, the content and type of the prediction results by model B2 do not limit this exemplary embodiment, but examples include: prediction results regarding the physical properties of a polymer material composed of one or more structural units (molecules) that constitute the monomer indicated by the monomer information MI, and physical properties of a polymer material composed of one or more structural units (molecules) that constitute the polymer chain indicated by the chain information CI.

[0081] 7, the prediction result (inference result) by the prediction process PB2 is provided to the update unit 122. The update unit 122 executes update process P122, which uses the prediction result by the prediction process PB2 and the reference result to update one or more parameters of the prediction model (at least one of model B1 and model B2).

[0082] As an example, the update unit 122 executes an update process P122 that updates one or more parameters of the prediction model so as to reduce the difference between: the prediction result by the prediction process PB2; and the correct label associated with at least one of the monomer information MI and the chain information CI.

[0083] Here, the correct label is, as an example, a correct label associated with the monomer information MI or the chain information CI, and is a correct label regarding the physical properties of the polymer material composed of the monomer indicated by the monomer information MI or one or more structural units (molecules) that constitute the polymer chain indicated by the chain information CI, but this does not limit this exemplary embodiment.

[0084] In addition, the update process P122 may include a process of updating one or more parameters of the prediction model (at least one of model B1 and model B2) by referring to a loss function expressed using the theoretical formula indicated by the theoretical formula information TFI.

[0085] As an example, the above-mentioned conversion process PB1, prediction process PB2, and update process P122 are repeatedly executed until the prediction model satisfies the desired properties (for example, until one or more parameters included in the prediction model satisfy specified convergence conditions).

[0086] The information processing device 1A according to this example includes an acquisition unit 11 that acquires monomer information and chain information, and a learning unit 12 that executes a learning process for a prediction model that makes predictions about a polymer material including one or more structural units that constitute one or more monomers, and the prediction model can also be expressed as including: a first model (model B1) to which the monomer information acquired by the acquisition unit 11 is input; and a second model (model B2) to which the output of the first model (model B1) and the chain information acquired by the acquisition unit 11 are input, the second model incorporating an equation formulated based on the laws of physics.

[0087] (Processing Example C) FIG. 8 is a diagram showing examples of processes executed by the learning unit 12 according to processing example C, and examples of information referenced or generated in each process.

[0088] As shown in Figure 8, in this example, the prediction unit 121 provided in the learning unit 12 executes: - a conversion process PC1 to oligomers referring to monomer information MI; - a conversion process PC2 to intermolecular information referring to chain information CI; - a theoretical formula selection process PC3 referring to at least one of the conversion results by the conversion process PC1 and the conversion results by the conversion process PC2 and theoretical formula information TFI; and - a prediction process PC4 using model C referring to the theoretical formula selected by the theoretical formula selection process PC3.

[0089] (Conversion process PC1) In the conversion process PC1, the prediction unit 121 generates oligomer information, which is information about an oligomer containing one or more constituent units (molecules) that constitute the monomer, from the molecular structure of one or more monomers contained in the monomer information MI.

[0090] In the conversion process PC2, the prediction unit 121 generates intermolecular information from information on at least one of the polymer chain, chain length, density, and molecular weight distribution indicated by the chain information CI. Here, the intermolecular information includes, for example, information on at least one of the interchain distance and orientation.

[0091] (Theoretical formula selection process PC3) In the theoretical formula selection process PC, the prediction unit 121 refers to at least one of the following to select a theoretical formula to be used in the prediction process PC4, which will be described later, from multiple candidate theoretical formulas included in the theoretical formula information TFI.

[0092] (Prediction Process PC4) In prediction process PC4, the prediction unit 121 generates a prediction result (inference result) by inputting at least one of the following into model C: monomer information MI; oligomer information generated from the monomer information MI by conversion process PC1; chain information CI; intermolecular information generated from the chain information CI by conversion process PC2; and physical property information PPI. Here, model C is an example of a prediction model defined by the above-mentioned model information MD. Furthermore, model C may or may not be a model incorporating the theoretical formula indicated by the theoretical formula information TFI.

[0093] The content and type of the prediction results by model C do not limit this exemplary embodiment, but examples include: prediction results regarding the physical properties of a polymeric material constituted by one or more structural units (molecules) that constitute a monomer indicated by the monomer information MI; physical properties of a polymeric material constituted by one or more structural units (molecules) that constitute a polymer chain indicated by the chain information CI; prediction results regarding the physical properties of a polymeric material constituted by one or more structural units (molecules) that constitute an oligomer indicated by the oligomer information; and prediction results regarding the physical properties of a polymeric material defined by information regarding at least one of the inter-chain distance and orientation indicated by the intermolecular information.

[0094] Thus, the learning process performed by the learning unit 12 in this example includes: a conversion process PC1 (first generation process) that generates oligomer information related to one or more oligomers by referring to the monomer information MI; and a conversion process PC2 (second generation process) that generates intermolecular information, which is information related to molecules, by referring to the chain information CI.

[0095] 8, the prediction result (inference result) by the prediction process PC4 is provided to the update unit 122. The update unit 122 executes update process P122, which uses the prediction result by the prediction process PC4 and the reference result to update one or more parameters of the prediction model (model C).

[0096] As an example, the update unit 122 executes an update process P122 that updates one or more parameters of the prediction model so as to reduce the difference between: the prediction result by the prediction process PC4; and the correct label associated with at least one of the monomer information MI, chain information CI, oligomer information, and intermolecular information.

[0097] Here, as an example, the correct label is a correct label associated with at least one of the monomer information MI, chain information CI, oligomer information, and intermolecular information, and is a correct label regarding: - the physical properties of a polymer material constituted by one or more structural units (molecules) that constitute at least one of the monomer indicated by the monomer information MI, the polymer chain indicated by the chain information CI, and the oligomer indicated by the oligomer information, or - the physical properties of a polymer material that are determined by the inter-chain distance and orientation indicated by the intermolecular information, but this does not limit the present exemplary embodiment.

[0098] In this example, the update process P122 updates one or more parameters of the prediction model (model C) by referring to a loss function expressed using the theoretical formula indicated by the theoretical formula information TFI.

[0099] As an example, in the update process P122, the loss function L L = L ｄａｔａ (d, td) + L ｐｈｙｓ (d) may be referred to. Here, the first term on the right side, L ｄａｔａ (d, td) is a loss function that quantitatively expresses the difference between the prediction result d by the prediction model and the correct label td to be compared with the prediction result. ｐｈｙｓ (d) is a loss function that quantitatively expresses the degree to which the prediction result d by the model satisfies the physical law (the theoretical formula described above). As an example, when the theoretical formula can be expressed as z = f(x, y) using variables x, y, z and a function f, and the prediction results by the model for the variables x, y, and z are xp, yp, and zp, L ｐｈｙｓ (d) is an example of L ｐｈｙｓ (d) = (zp-f(xp, yp)) ２In other words, L ｐｈｙｓ (d) may be expressed as the degree to which the predicted result d by the prediction model satisfies the theoretical formula (or the degree to which the theoretical formula is not satisfied), although this example does not limit the present exemplary embodiment.

[0100] As an example, the above-mentioned conversion process PA1, prediction process PA2, and update process P122 are repeatedly executed until the prediction model satisfies the desired properties (for example, until one or more parameters included in the prediction model satisfy predetermined convergence conditions).

[0101] The information processing device 1A according to this example includes an acquisition unit 11 that acquires monomer information and chain information, and a learning unit 12 that executes a learning process for a prediction model that makes predictions about a polymer material including one or more structural units that constitute one or more monomers, and the learning process can also be expressed as including a process of updating one or more parameters of the prediction model (model C) by referring to a loss function expressed using an equation formulated based on the laws of physics.

[0102] (Example of Theoretical Formula) The following is a specific example of the theoretical formula referenced in the learning process by the learning unit 12. However, it is clear that the following example is merely an example and does not limit the present exemplary embodiment.

[0103] (Example 1: Theoretical Formula for Diffusion Coefficient of Polymer Chain) In the learning process by the learning unit 12, the following theoretical formula for the diffusion coefficient of polymer chain, or information obtained using the formula, may be referenced.

[0104] L ｐｈｙｓ = a・D D = (Mw) －２．３ Here, a is a parameter (coefficient) to be updated in the learning process or the preliminary fitting process by the learning unit 12, D represents diffusion, Mw represents the molecular weight of the polymer chain, and "·" represents a product.

[0105] In the above-described processing examples A to C, the Lｐｈｙｓ Alternatively, a learning process may be performed using a prediction model incorporating the L ｐｈｙｓ A learning process may be performed with reference to a loss function including:

[0106] (Example 2: Entropy Factor Related to Mechanical Stiffness) In the learning process by the learning unit 12, the following theoretical formula related to the entropy factor related to mechanical stiffness or information obtained using the formula may be referenced.

[0107] L ｐｈｙｓ = N log N where N indicates the number of polymers (the number of polymer chains) contained in the system, and as an example, it is given by N ~ Msys / Mw using the molecular weight of the system Msys and the molecular weight of the polymer chain Mw. ｐｈｙｓ is a quantitative expression of polymer entanglement and is formulated based on statistical mechanics.

[0108] In the above-described processing examples A to C, the L ｐｈｙｓ Alternatively, a learning process may be performed using a prediction model incorporating the L ｐｈｙｓ A learning process may be performed with reference to a loss function including:

[0109] 9 shows an example of a display by the input / output unit 17A included in the information processing device 1 A. As shown in Fig. 9, the display included in the input / output unit 17A may be configured as follows: in a display region RR1, to display monomer information referenced in the learning process by the learning unit 12, in a display region RR2, to display chain information referenced in the learning process by the learning unit 12, and in a display region RR3, to display physical property information referenced in the learning process by the learning unit 12 or physical property information indicating a prediction result generated by a prediction model in the learning process by the learning unit 12.

[0110] Here, display region RR1 may function as a reception area for the user to input monomer information. Display region RR2 may also function as a reception area for the user to input chain information. Display region RR3 may also function as a reception area for the user to input desired physical property information. In this configuration, a learning process is executed with reference to the physical property information received via display region RR3. As an example, in the above-described processing examples A to C, a configuration may be adopted in which one or more theoretical formulas are selected from multiple candidates with reference to the physical property information, and a learning process is executed using the selected one or more theoretical formulas.

[0111] With this configuration, it is possible to preferably execute the learning process regarding the physical property information desired by the user.

[0112] (Effects of Information Processing Device 1A) As described above, the information processing device 1A is configured to acquire monomer information, which is information about one or more monomers, and execute a learning process for a prediction model that makes predictions about a polymer material containing one or more structural units that make up one or more monomers, by referring to at least the acquired monomer information and an equation formulated based on physical laws, or information obtained using the equation.

[0113] In this way, the information processing device 1A performs a learning process that at least references an equation formulated based on the laws of physics or information obtained using the equation, thereby making it possible to effectively construct a predictive model that makes predictions about polymer materials while keeping computational costs down.

[0114] (Information processing device 2A) Fig. 10 is a block diagram showing the configuration of an information processing device 2A according to this exemplary embodiment. As shown in Fig. 10, the information processing device 2A includes a control unit 20A, a storage unit 15A, a communication unit 16A, and an input / output unit 17A. In the following, the matters described in relation to the information processing device 1A will not be described again.

[0115] 10, the control unit 20A includes an acquisition unit 21 and a prediction unit 22. The acquisition unit 21 acquires monomer information, which is information about one or more monomers. Here, the monomer information is, for example, information including at least one molecular structure of the monomer.

[0116] The prediction unit 22 generates a prediction result by inputting at least the monomer information acquired by the acquisition unit 21 into a prediction model that makes predictions regarding a polymeric material that includes one or more structural units that constitute one or more monomers.

[0117] Here, the prediction model is a model constructed by a learning process that refers to at least: monomer information for learning; and an equation formulated based on physical laws, or information obtained using the equation. As an example, the prediction model may be a prediction model trained by the information processing device 1A described above.

[0118] (Effects of information processing device 2A) As described above, the information processing device 2A acquires monomer information, which is information about one or more monomers, and generates a prediction result by inputting at least the acquired monomer information into a prediction model that makes predictions about a polymer material containing one or more structural units that constitute one or more monomers, and the prediction model is constructed by a learning process that at least references the monomer information for learning, and an equation formulated based on physical laws, or information obtained using the equation.

[0119] In this way, the information processing device 2A generates prediction results using a formula based on the laws of physics or a prediction model constructed by a learning process that at least references information obtained using the formula, so that predictions regarding polymer materials can be made effectively while keeping calculation costs down.

[0120] [Example of implementation by software] Some or all of the functions of the information processing devices 1, 2, 1A, 2A (hereinafter also referred to as "each of the above devices") may be implemented by hardware such as an integrated circuit (IC chip), or by software.

[0121] 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 11. Figure 11 is a block diagram showing the hardware configuration of computer C that functions as each of the above devices.

[0122] 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 function 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.

[0123] The processor C1 may be, for example, a central processing unit (CPU), a graphics 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.

[0124] 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, a mouse, a display, and a printer.

[0125] The program P can also 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.

[0126] Furthermore, the functions of each of the devices may be realized by a single processor provided in a single computer, by multiple processors provided in a single computer working in cooperation, or by multiple processors provided in each of multiple computers working in cooperation. Furthermore, the programs for causing each of the devices to realize the functions may be stored in a single memory provided in a single computer, or may be distributed and stored in multiple memories provided in a single computer, or may be distributed and stored in multiple memories provided in each of multiple computers.

[0127] [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.

[0128] (Appendix A1) An information processing device comprising: an acquisition means for acquiring monomer information, which is information relating to one or more monomers; and a learning means for executing a learning process for a prediction model that makes predictions regarding a polymer material including one or more structural units that constitute one or more monomers, by referring to at least the monomer information acquired by the acquisition means, and an equation formulated based on physical laws, or information obtained using the equation.

[0129] (Appendix A2) The information processing device according to Appendix A1, wherein the acquisition means further acquires physical property information that is information relating to a physical property of interest, and the learning means executes the learning process by further referring to the physical property information.

[0130] (Appendix A3) The information processing device according to Appendix A2, wherein the monomer information includes at least one molecular structure of a monomer.

[0131] (Appendix A4) The information processing device according to Appendix A2 or A3, wherein the acquisition means further acquires chain information that is information relating to a polymer chain constituted by the one or more structural units, and the learning means executes the learning process by further referring to the chain information.

[0132] (Appendix A5) The information processing device according to Appendix A4, wherein the chain information includes information on at least one of a chain length of the polymer chain, a density of the polymer chain, and a molecular weight distribution of the polymer chain.

[0133] (Supplementary Note A6) The information processing device according to Supplementary Note A4 or A5, wherein the learning process includes a selection process of selecting the formula from one or more candidates for the formula by referring to at least one of the physical property information and the chain information.

[0134] (Appendix A7) The information processing device according to any one of Appendices A4 to A6, wherein the learning process includes a process of inputting the monomer information and information obtained by inputting the chain information into the formula into the prediction model.

[0135] (Appendix A8) The information processing device according to any one of Appendices A4 to A6, wherein the prediction model includes: a first model to which the monomer information is input; and a second model to which the output of the first model and the chain information are input, the second model incorporating the formula.

[0136] (Appendix A9) The information processing device according to Appendix A6, wherein the learning process includes: a first generation process that generates oligomer information relating to one or more oligomers by referring to the monomer information; and a second generation process that generates intermolecular information that is information relating to molecules by referring to the chain information; and the selection process further refers to the intermolecular information.

[0137] (Appendix A10) The information processing device according to Appendix A9, wherein the intermolecular information includes information relating to at least one of an interchain distance and an orientation.

[0138] (Supplementary Note A11) The information processing device according to any one of Supplementary Notes A1 to A10, wherein the learning process includes a process of updating one or more parameters included in the prediction model by referring to a loss function expressed using the formula.

[0139] [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.

[0140] (Appendix B1) An information processing method including: an acquisition step of acquiring monomer information, which is information relating to one or more monomers; and a learning step of executing a learning process for a prediction model that makes predictions regarding a polymer material including one or more structural units that constitute one or more monomers, by referring to at least the monomer information acquired in the acquisition step, and an equation formulated based on physical laws, or information obtained using the equation.

[0141] (Supplementary Note B2) The information processing method according to Supplementary Note B1, wherein the acquiring step further acquires physical property information that is information relating to a physical property of interest, and the learning step executes the learning process by further referring to the physical property information.

[0142] (Appendix B3) The information processing method according to Appendix B2, wherein the monomer information includes at least one molecular structure of a monomer.

[0143] (Appendix B4) The information processing method according to Appendix B2 or B3, wherein the acquiring step further acquires chain information, which is information relating to a polymer chain constituted by the one or more structural units, and the learning step executes the learning process by further referring to the chain information.

[0144] (Appendix B5) The information processing method according to Appendix B4, wherein the chain information includes information on at least one of the chain length of the polymer chain, the density of the polymer chain, and the molecular weight distribution of the polymer chain.

[0145] (Appendix B6) The information processing method according to Appendix B4 or B5, wherein the learning process includes a selection process of selecting the formula from one or more candidates for the formula by referring to at least one of the physical property information and the chain information.

[0146] (Appendix B7) The information processing method according to any one of Appendices B4 to B6, wherein the learning process includes a process of inputting the monomer information and information obtained by inputting the chain information into the formula into the prediction model.

[0147] (Appendix B8) The information processing method according to any one of Appendices B4 to B6, wherein the prediction model includes: a first model to which the monomer information is input; and a second model to which the output of the first model and the chain information are input, the second model incorporating the formula.

[0148] (Appendix B9) The information processing method according to Appendix B6, wherein the learning process includes: a first generation process that generates oligomer information relating to one or more oligomers by referring to the monomer information; and a second generation process that generates intermolecular information, which is information relating to molecules, by referring to the chain information; and the selection process further refers to the intermolecular information.

[0149] (Appendix B10) The information processing method according to Appendix B9, wherein the intermolecular information includes information relating to at least one of an inter-chain distance and an orientation.

[0150] (Appendix B11) The information processing method according to any one of Appendices B1 to B10, wherein the learning process includes a process of updating one or more parameters included in the prediction model by referring to a loss function expressed using the formula.

[0151] [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.

[0152] (Appendix C1) A program that causes a computer to function as an information processing device, causing the computer to function as: an acquisition means that acquires monomer information, which is information about one or more monomers; and a learning means that executes a learning process for a prediction model that makes predictions about a polymer material including one or more structural units that constitute one or more monomers, by referring to at least the monomer information acquired by the acquisition means, and an equation formulated based on physical laws, or information obtained using the equation.

[0153] (Supplementary Note C2) The information processing program according to Supplementary Note C1, wherein the acquisition means further acquires physical property information that is information relating to a physical property of interest, and the learning means executes the learning process by further referring to the physical property information.

[0154] (Appendix C3) The information processing program according to Appendix C2, wherein the monomer information includes at least one molecular structure of a monomer.

[0155] (Appendix C4) The information processing program according to Appendix C2 or C3, wherein the acquisition means further acquires chain information, which is information relating to a polymer chain constituted by the one or more structural units; and the learning means executes the learning process by further referring to the chain information.

[0156] (Appendix C5) The information processing program according to Appendix C4, wherein the chain information includes information on at least one of a chain length of a polymer chain, a density of a polymer chain, and a molecular weight distribution of a polymer chain.

[0157] (Appendix C6) The information processing program according to Appendix C4 or C5, wherein the learning process includes a selection process of selecting the formula from one or more candidates for the formula by referring to at least one of the physical property information and the chain information.

[0158] (Appendix C7) The information processing program according to any one of Appendices C4 to C6, wherein the learning process includes a process of inputting the monomer information and information obtained by inputting the chain information into the formula into the prediction model.

[0159] (Appendix C8) The information processing program according to any one of Appendices C4 to C6, wherein the prediction model includes: a first model to which the monomer information is input; and a second model to which the output of the first model and the chain information are input, the second model incorporating the formula.

[0160] (Appendix C9) The information processing program according to Appendix C6, wherein the learning process includes: a first generation process that generates oligomer information relating to one or more oligomers by referring to the monomer information; and a second generation process that generates intermolecular information that is information relating to molecules by referring to the chain information; and the selection process further refers to the intermolecular information.

[0161] (Appendix C10) The information processing program according to Appendix C9, wherein the intermolecular information includes information on at least one of an inter-chain distance and an orientation.

[0162] (Appendix C11) The information processing program according to any one of Appendices C1 to C10, wherein the learning process includes a process of updating one or more parameters included in the prediction model by referring to a loss function expressed using the formula.

[0163] [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.

[0164] (Appendix D1) An information processing device comprising at least one processor, wherein the at least one processor executes an acquisition process for acquiring monomer information, which is information relating to one or more monomers, and further executes a learning process for a prediction model that makes predictions regarding a polymer material including one or more structural units that constitute one or more monomers, by referring at least to the monomer information acquired by the acquisition process, and an equation formulated based on the laws of physics, or information obtained using the equation.

[0165] 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.

[0166] (Appendix D2) The information processing device according to Appendix D1, wherein in the acquisition process, the at least one processor further acquires physical property information that is information relating to a physical property of interest, and in the learning process, the at least one processor further refers to the physical property information.

[0167] (Appendix D3) The information processing device according to Appendix D2, wherein the monomer information includes at least one molecular structure of a monomer.

[0168] (Appendix D4) The information processing device according to Appendix D2 or D3, wherein in the acquisition process, the at least one processor further acquires chain information, which is information relating to a polymer chain composed of the one or more structural units; and in the learning process, the at least one processor further refers to the chain information.

[0169] (Appendix D5) The information processing device according to appendix D4, wherein the chain information includes information on at least one of a chain length of the polymer chain, a density of the polymer chain, and a molecular weight distribution of the polymer chain.

[0170] (Appendix D6) The information processing device according to appendix D4 or D5, wherein in the learning process, the at least one processor executes a selection process to select the formula from one or more candidates for the formula by referring to at least one of the physical property information and the chain information.

[0171] (Appendix D7) The information processing device according to any one of Appendices D4 to D6, wherein the learning process includes a process of inputting the monomer information and information obtained by inputting the chain information into the formula into the prediction model.

[0172] (Appendix D8) The information processing device according to any one of Appendices D4 to D6, wherein the prediction model includes: a first model to which the monomer information is input; and a second model to which the output of the first model and the chain information are input, the second model incorporating the formula.

[0173] (Appendix D9) The information processing device described in Appendix D6, wherein in the learning process, the at least one processor executes a first generation process that generates oligomer information regarding one or more oligomers by referring to the monomer information, and a second generation process that generates intermolecular information that is information regarding intermolecular relationships by referring to the chain information, and in the selection process, the at least one processor further refers to the intermolecular information.

[0174] (Appendix D10) The information processing device according to appendix D9, wherein the intermolecular information includes information relating to at least one of an inter-chain distance and an orientation.

[0175] (Supplementary Note D11) The information processing device according to any one of Supplementary Notes D1 to D10, wherein in the learning process, the at least one processor executes a process of updating one or more parameters included in the prediction model by referring to a loss function expressed using the formula.

[0176] [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.

[0177] (Appendix E1) A non-transitory recording medium having recorded thereon an information processing program that causes a computer to function as an information processing device, the information processing program causing the computer to execute an acquisition process that acquires monomer information, which is information about one or more monomers, and a learning process that executes a prediction model that makes predictions about polymeric materials containing one or more structural units that constitute one or more monomers, by referring to at least the monomer information acquired by the acquisition process, and an equation formulated based on the laws of physics, or information obtained using the equation.

[0178] 1, 2, 1A, 2A ... Information processing device 11, 21 ... Acquisition unit 12 ... Learning unit 22 ... Prediction unit

Claims

1. An information processing device comprising: an acquisition means for acquiring monomer information, which is information relating to one or more monomers; and a learning means for executing a learning process for a prediction model that makes predictions regarding polymeric materials containing one or more structural units that constitute one or more monomers, by referring at least to the monomer information acquired by the acquisition means, and an equation formulated based on the laws of physics, or information obtained using said equation.

2. The information processing device according to claim 1, wherein the acquisition means further acquires physical property information, which is information relating to a physical property of interest, and the learning means executes the learning process by further referring to the physical property information.

3. The information processing device according to claim 2, wherein the monomer information includes at least one molecular structure of the monomer.

4. The information processing device according to claim 2 or 3, wherein the acquisition means further acquires chain information, which is information relating to a polymer chain composed of the one or more structural units, and the learning means executes the learning process by further referring to the chain information.

5. The information processing device according to claim 4, wherein the chain information includes information on at least one of the chain length of the polymer chain, the density of the polymer chain, and the molecular weight distribution of the polymer chain.

6. An information processing device according to claim 4 or 5, wherein the learning process includes a selection process for selecting the formula from one or more candidates for the formula by referring to at least one of the physical property information and the chain information.

7. The information processing device according to any one of claims 4 to 6, wherein the learning process includes a process of inputting the monomer information and information obtained by inputting the chain information into the formula into the prediction model.

8. An information processing device according to any one of claims 4 to 6, wherein the prediction model includes: a first model to which the monomer information is input; and a second model to which the output of the first model and the chain information are input, the second model incorporating the formula.

9. The information processing device according to claim 6, wherein the learning process includes: a first generation process that generates oligomer information relating to one or more oligomers by referring to the monomer information; and a second generation process that generates intermolecular information, which is information relating to molecules, by referring to the chain information; and wherein the selection process further refers to the intermolecular information.

10. The information processing device according to claim 9, wherein the intermolecular information includes information relating to at least one of interchain distance and orientation.

11. An information processing device according to any one of claims 1 to 10, wherein the learning process includes a process of updating one or more parameters included in the prediction model by referring to a loss function expressed using the formula.

12. An information processing device comprising: an acquisition means for acquiring monomer information, which is information relating to one or more monomers; and a prediction means for generating prediction results by inputting at least the monomer information acquired by said acquisition means into a prediction model that makes predictions regarding polymeric materials containing one or more structural units that constitute one or more monomers, wherein said prediction model is a model constructed by a learning process that at least references the monomer information for learning, and an equation formulated based on physical laws, or information obtained using said equation.

13. An information processing method comprising: acquiring monomer information, which is information relating to one or more monomers; and executing a learning process for a predictive model that makes predictions regarding polymeric materials containing one or more structural units that constitute one or more monomers, by referring to at least the acquired monomer information and an equation formulated based on physical laws, or information obtained using said equation.

14. An information processing method comprising: acquiring monomer information, which is information relating to one or more monomers; and generating a prediction result by inputting at least the acquired monomer information into a prediction model that makes predictions regarding polymeric materials containing one or more structural units that constitute one or more monomers, wherein the prediction model is a model constructed by a learning process that at least references the monomer information for learning, and an equation formulated based on physical laws, or information obtained using said equation.

15. A program that causes a computer to function as an information processing device, said program causing the computer to execute an acquisition process to acquire monomer information, which is information relating to one or more monomers, and further execute a learning process for a prediction model that makes predictions regarding polymeric materials containing one or more structural units that constitute one or more monomers, by referring at least to the monomer information acquired by the acquisition process and an equation formulated based on physical laws, or information obtained using said equation.

16. A program that causes a computer to function as an information processing device, the program causing the computer to execute an acquisition process that acquires monomer information, which is information relating to one or more monomers, and a prediction process that generates a prediction result by inputting at least the monomer information acquired by the acquisition process into a prediction model that makes predictions regarding polymeric materials containing one or more structural units that constitute one or more monomers, the prediction model being a model constructed by a learning process that at least references the monomer information for learning and an equation formulated based on the laws of physics, or information obtained using said equation.

Images