Self-supervised learning programs, methods, and apparatus

Abstract

High-precision machine learning models are generated through self-supervised learning. [Solution] The self-supervised learning device generates data representing a second molecule in which each of the multiple atoms contained in the first molecule is replaced with any of the atoms contained in the first molecule. The data representing the second molecule is input into a machine learning model that makes predictions about molecular structure to obtain prediction results. Based on the comparison results between the correct data corresponding to the first molecule and the prediction results, the parameters of the machine learning model are updated.

Description

Technical Field

[0001] The disclosed technology relates to a self-supervised learning program, a self-supervised learning method, and a self-supervised learning device.

Background Art

[0002] Material discovery, which predicts the properties of material structures and discovers or optimizes new materials suitable for various applications, is carried out in various fields and industries and is essential for the development of technology. Specific examples of the objectives of material discovery include the search for high-performance materials for aircraft, materials for perovskite solar cells, new catalysts that efficiently promote chemical reactions, materials for biodegradable plastics, and highly biocompatible medical alloys.

[0003] In material discovery, the properties of material structures are predicted through experiments or simulations. As simulations, for example, molecular dynamics (MD) simulations that perform simulations based on Newton's equations of motion to obtain the forces and energies of atoms and molecules are used. MD simulations have the advantage of being able to calculate the time evolution of a system and having a relatively low computational cost. Also, density functional theory (DFT) calculations that perform simulations based on first principles calculations based on quantum mechanics to obtain the forces and energies of atoms and molecules are used. DFT calculations generally calculate the ground state and have a relatively high computational cost.

[0004] Recently, material discovery is sometimes carried out using Materials Informatics (MI), which utilizes large amounts of data. MI is a field in materials science that uses AI (Artificial Intelligence) and other technologies to clarify the properties of materials and discover new materials. MI provides methods for analyzing large amounts of data and predicting the properties and behavior of materials. One example of an MI technique is high-throughput screening, which rapidly selects materials that meet certain criteria for a given problem setting. High-throughput screening is used to quickly evaluate a large number of material candidates using AI and other technologies to identify the optimal material. There is also a method that uses AI surrogate models to directly predict the results obtained from MD simulations and DFT calculations. Methods using AI surrogate models are used to replace computationally expensive simulations and can significantly reduce computation time.

[0005] Thus, in recent years, methods for rapidly predicting material properties using AI models have been developed. Various AI models have been proposed for predicting material properties, such as SchNet, PaiNN, GemNet-OC, and EquiformerV2.

[0006] To train an AI model, correct label data is necessary. For example, in machine learning in the field of materials, the input data is a graph of the chemical structure considering the atomic species and the position of each atom, and the labels are force, energy, etc. This label data related to material structures can be obtained through experiments or simulations. However, experiments require advanced equipment and expertise, and are time-consuming and expensive. Simulations, especially DFT, are computationally expensive and require large computing resources. In other words, both experiments and simulations are costly to conduct, making it difficult to collect large amounts of label data.

[0007] Therefore, there is a method called self-supervised learning, in which an AI model is pre-trained using a large amount of unlabeled data, and then trained using a small amount of labeled data. This method is also used in the field of materials discovery.

[0008] For example, an information processing system has been proposed that generates a trained model by performing machine learning based on sequence information representing sequences that include part or all of the sequences of multiple antigen-binding molecules or proteins. Based on the trained model, this system generates virtual sequence information in which at least one of the constituent units that make up the sequence represented by the sequence information input to the trained model has been mutated.

[0009] Furthermore, a method has been proposed for pre-training a graph neural network using, for example, an untagged compound dataset. In this method, one of the tagged compound datasets tags one of the compound's properties. This method further obtains a compound property prediction model by performing multi-task training on a graph neural network pre-trained using multiple tagged compound datasets. The compound property prediction model is used to predict various properties of a compound. [Prior art documents] [Patent Documents]

[0010] [Patent Document 1] International Publication No. 2020 / 246617 [Patent Document 2] Japanese Patent Publication No. 2022-163051 [Overview of the Initiative] [Problems that the invention aims to solve]

[0011] Traditional self-supervised learning methods for training machine learning models still have room for improvement in terms of accuracy.

[0012] One aspect of the disclosure technology is that it aims to generate highly accurate machine learning models through self-supervised learning. [Means for solving the problem]

[0013] In one embodiment, the disclosed technology generates data representing a second molecule in which each of a plurality of atoms contained in a first molecule is replaced with any of the atoms contained in the first molecule. The disclosed technology then inputs the data representing the second molecule into a machine learning model that makes predictions about molecular structure to obtain prediction results, and updates the parameters of the machine learning model based on the comparison result between the prediction results and the ground truth data corresponding to the first molecule. [Effects of the Invention]

[0014] One aspect of this approach is that it allows for the generation of highly accurate machine learning models through self-supervised learning. [Brief explanation of the drawing]

[0015] [Figure 1] This is a diagram illustrating neural networks (NNs) used to predict the energy of molecules. [Figure 2] This is a diagram to explain transfer learning. [Figure 3] This is a diagram to explain Reference Method 1. [Figure 4] This is a diagram to explain Reference Method 2. [Figure 5] This is a diagram to explain fine tuning. [Figure 6] This diagram illustrates the problems with the reference method. [Figure 7] This is a diagram illustrating the outline of this embodiment. [Figure 8] This is a functional block diagram of a self-supervised learning device according to the first to fourth embodiments. [Figure 9] This is a diagram illustrating the processing of the generation unit according to the first embodiment. [Figure 10]It is a block diagram showing a schematic configuration of a computer that functions as a self-supervised learning device according to the first to fourth embodiments. [Figure 11] It is a flowchart showing an example of the self-supervised learning process according to the first embodiment. [Figure 12] It is a diagram for explaining another process of the generation unit according to the first embodiment. [Figure 13] It is a diagram for explaining the process of the generation unit according to the second embodiment. [Figure 14] It is a flowchart showing an example of the self-supervised learning process according to the second embodiment. [Figure 15] It is a flowchart showing an example of the self-supervised learning process according to the third embodiment. [Figure 16] It is a flowchart showing an example of the self-supervised learning process according to the fourth embodiment.

Embodiments for Carrying Out the Invention

[0016] Hereinafter, an example of an embodiment according to the disclosed technology will be described with reference to the drawings.

[0017] Before explaining the details of the embodiment, self-supervised learning of a machine learning model for predicting the properties of materials will be described.

[0018] For example, as shown in FIG. 1, a neural network (hereinafter also referred to as "NN") that predicts the energy of a molecule based on the molecular structure data is trained by supervised learning using labeled data consisting of "structure" and "energy". That is, the molecular structure data is input into the NN, and the parameters of the NN are updated by backpropagating the difference between the predicted energy of the molecule, which is the output from the NN, and the correct energy. This correct energy is calculated by a simulation such as DFT, for example. Since the calculation time of the energy of a molecule by DFT is very long, it is difficult to collect a large amount of labeled data for supervised learning.

[0019] Therefore, there is a technique called self-supervised learning, which generates correct answers (labels) for a task from unlabeled data and performs supervised learning using the generated labels. Generally, self-supervised learning is applied to pre-training in transfer learning. As shown in Figure 2, transfer learning pre-trains a neural network (NN) using self-supervised learning with unlabeled data. The NN includes a feature extraction unit that extracts features from the input data and an output unit that corresponds to the task. The output unit used during pre-training is a pre-training unit. In pre-training, machine learning is performed using a large amount of data similar to the data domain that is actually to be specialized. By using a large amount of data to train the NN on the essential features of the input molecular data, the initial state of the NN can be improved.

[0020] Furthermore, if the tasks for pre-training and fine-tuning (FT) are different, the output layer is modified for FT. For example, if the pre-training task is 3-class classification but the fine-tuning task is 4-class classification, the output layer is changed to a fully connected layer for 4-class classification. After that, the neural network (NN) is trained using supervised learning with labeled data for fine-tuning. The feature extraction unit at this stage is the same as the one pre-trained. That is, the weights of the pre-trained NN are used as initial values, and machine learning is performed on the actual data. By applying self-supervised learning as pre-training, the knowledge gained from pre-training can be utilized, and even when there is little labeled data needed for fine-tuning (supervised learning) after pre-training, the accuracy of the NN can be improved.

[0021] Two example methods are presented for self-supervised learning of neural networks (NNs) used to predict material properties. Method 1, as shown in Figure 3, prepares molecular data containing information on the atomic species and 3D position coordinates of each of the multiple atoms constituting the molecule. In the example in Figure 3, each atom is represented by a circle, with α, β, etc., inside the circle representing the atomic species.

[0022] In Reference Method 1, a predetermined proportion of atoms in the prepared molecular data are replaced with other atoms. In the example in Figure 3, the shaded circles represent the atoms that have been replaced from the original molecular data. Then, the neural network (NN) is asked to predict whether each atom in the substituted molecular data has been replaced from each atom in the original molecular data. In other words, Reference Method 1 sets up an output unit (for pre-training) for performing a binary classification task. Since Reference Method 1 can generate ground truth data from the difference between the original molecular data and the substituted molecular data, self-supervised learning is possible. In the example in Figure 3, atoms that have not been replaced are represented by "0" and atoms that have been replaced are represented by "1" in both the ground truth data and the prediction results.

[0023] Reference Method 2, as shown in Figure 4, replaces a predetermined proportion of atoms in the prepared molecular data with other atoms, and then has the neural network predict the original atom species before the substitution for each atom included in the substituted molecular data. In other words, Reference Method 2 sets up an output unit (for pre-training) for performing a multi-class classification task. In Reference Method 2, the original molecular data becomes the ground truth data, making self-supervised learning possible.

[0024] In the examples in Figures 3 and 4, when referring to the ground truth data, the predictions made by the neural network (NN) are incorrect for the atoms indicated by the shaded areas. The NN parameters are updated so that these predictions match the ground truth data.

[0025] In both Reference Methods 1 and 2, as shown in Figure 5, the feature extraction unit of the NN, pre-trained by self-supervised learning, is used as the initial value, and fine-tuning is performed with the actual training data. During the fine-tuning stage, it is assumed that the feature extraction unit of the NN has learned essential features such as the structure of atoms and molecules. Furthermore, since the output unit of the NN during pre-training is for predicting binary or multi-class classification tasks, in fine-tuning, an output unit is set according to the actual task, such as predicting force or energy.

[0026] In reference methods 1 and 2, as shown in Figure 6, the atomic species of the atoms in the original molecular data (shaded circles in Figure 6) are replaced with atomic species randomly selected from a pre-defined group of atomic species. In this case, since the replacement is with a unique atom that has a different tendency from other atoms in the molecule, it is easy to identify which atom has been replaced. Furthermore, it is easy to determine the original atomic species of the replaced atom from the atomic species of the atoms surrounding the unique atom (thick-lined circles in Figure 6). As a result, the neural network can predict the correct answer without learning the essential characteristics of the entire molecular structure, which reduces the accuracy of the neural network as a tool for predicting material properties.

[0027] Therefore, in this embodiment, as shown in Figure 7, molecular data for self-supervised learning is generated without creating unique atoms within the molecule by shuffling the atoms contained in the original molecular data. In the lower part of Figure 7, the shaded atoms represent that atomic species α in the original molecular data has been replaced by atomic species β, and the atoms with thick lines represent that atomic species β in the original molecular data has been replaced by atomic species α. In Figures 6 and 7, α and β represent atomic species such as atomic number and element symbol. The same applies to α, β, and γ in the following figures. The self-supervised learning apparatus according to each embodiment will be described in detail below.

[0028] <First Embodiment> As shown in Figure 8, the self-supervised learning device 10 according to the first embodiment functionally includes a generation unit 12, an acquisition unit 14, and an update unit 16. A neural network (NN) 24 is stored in a predetermined memory area of ​​the self-supervised learning device 10. The NN 24 is an example of a "machine learning model" in the disclosed technology.

[0029] As shown in Figure 8, the self-supervised learning device 10 receives first molecular data. The first molecular data includes, for example, information on the atomic species and three-dimensional position coordinates of each atom contained in the first molecule. The first molecular data may also include information indicating the bonding relationships between atoms.

[0030] The generation unit 12 acquires the first molecular data input to the self-supervised learning device 10. The generation unit 12 generates second molecular data representing a second molecule in which each of the multiple atoms contained in the first molecule is replaced with one of the atoms contained in the first molecule. Specifically, the generation unit 12 generates the second molecular data by shuffling the atoms contained in the first molecule in a complete permutation.

[0031] For example, the generation unit 12 stores the first molecular data as an atomic species list and a coordinate list. The atomic species list is a list of the length (number of elements) corresponding to the number of atoms contained in the first molecule, and the value of each element is a symbol representing the atomic species, such as the atomic number or element symbol. The coordinate list is a list of the size (number of atoms × 3) representing the three-dimensional position coordinates of each atom contained in the first molecule. The three-dimensional position coordinates of the atom corresponding to the i-th element of the atomic species list are stored in the i-th element of the coordinate list.

[0032] As shown in Figure 9A, the generation unit 12 assigns indices 0, 1, 2..., N-1 (where N is the number of atoms in the first molecule) to the original atom list in order from the beginning. In Figure 9, the numbers written next to each element in the list are the indices corresponding to that element. The generation unit 12 generates random integers from 0 to N-1, rearranging the integers assigned as indices to the original atom list in a complete permutation and associating them in order from the beginning of the substituted atom list. The substituted atom list is a list with the same number of elements as the original atom list, and at this stage, each element is empty. As shown in Figure 9B, the generation unit 12 generates a shuffled atom list by storing the atom species stored in the element of the original atom list that corresponds to the index of that element in each element of the substituted atom list. The generation unit 12 combines the shuffled atom list with the coordinate list of the first molecule data to generate the second molecule data.

[0033] NN24 includes a feature extraction unit that extracts features representing the molecular structure of a second molecule from second molecular data, and an output unit that outputs prediction results according to the pre-training task based on the features extracted by the feature extraction unit. The pre-training task may be a binary classification task, as shown in Figure 3, which predicts whether each atom in the second molecule has been replaced by each atom in the first molecule. Alternatively, the pre-training task may be a multi-class classification task, as shown in Figure 4, which predicts the original atom before replacement for each atom in the second molecular data.

[0034] The acquisition unit 14 inputs the second molecular data generated by the generation unit 12 to the NN24 and acquires the prediction results output from the NN24. In the case of the binary classification task described above, the prediction results are information indicating whether or not each atom in the second molecule has been substituted from the first molecule, and in the case of the multi-class classification task, they are information indicating the original atomic species of each atom in the second molecule before substitution.

[0035] The update unit 16 updates the parameters of the NN24 based on the comparison result between the ground truth data corresponding to the first molecule and the prediction result obtained by the acquisition unit 14. Specifically, in the case of the binary classification task described above, the update unit 16 generates ground truth data in the first molecular data and the second molecular data where corresponding atoms with the same 3D position coordinates are [0] if they are the same atomic species, and [1] if they are different. In the case of the multi-class classification task described above, the update unit 16 uses the first molecular data as the ground truth data. The update unit 16 then backpropagates the difference between the prediction result and the ground truth data to the NN24 and updates the parameters of the NN24.

[0036] The update unit 16 repeatedly updates the parameters of NN24 until a predetermined termination condition is met. The predetermined termination condition may be, for example, when the number of iterations reaches a predetermined number, when the difference between the prediction result and the ground truth data falls below a predetermined value, or when the difference between the previous difference and the current difference falls below a predetermined value (when the difference converges).

[0037] The self-supervised learning device 10 may be implemented, for example, by the computer 40 shown in Figure 10. The computer 40 comprises a CPU (Central Processing Unit) 41, a GPU (Graphics Processing Unit) 42, a memory 43 as a temporary storage area, and a non-volatile storage device 44. The computer 40 also comprises input / output devices 45 such as input devices and display devices, and an R / W (Read / Write) device 46 that controls the reading and writing of data to and from the storage medium 49. The computer 40 also comprises a communication interface 47 that connects to a network such as the Internet. The CPU 41, GPU 42, memory 43, storage device 44, input / output devices 45, R / W device 46, and communication interface 47 are connected to each other via a bus 48.

[0038] The storage device 44 is, for example, an HDD (Hard Disk Drive), an SSD (Solid State Drive), or flash memory. The storage device 44 stores a self-supervised learning program 50 that causes the computer 40 to function as a self-supervised learning device 10. The self-supervised learning program 50 includes a generation process control instruction 52, an acquisition process control instruction 54, and an update process control instruction 56. The storage device 44 also has an information storage area 60 in which information constituting the NN24 is stored.

[0039] The CPU 41 reads the self-supervised learning program 50 from the storage device 44, loads it into memory 43, and sequentially executes the control instructions contained in the self-supervised learning program 50. The CPU 41 operates as the generation unit 12 shown in Figure 8 by executing the generation process control instruction 52. The CPU 41 also operates as the acquisition unit 14 shown in Figure 8 by executing the acquisition process control instruction 54. The CPU 41 also operates as the update unit 16 shown in Figure 8 by executing the update process control instruction 56. The CPU 41 also reads information from the information storage area 60 and loads NN24 into memory 43. As a result, the computer 40 that executed the self-supervised learning program 50 functions as a self-supervised learning device 10. The CPU 41 that executes the program is hardware. Also, part of the program may be executed by the GPU 42.

[0040] Furthermore, the functions realized by the self-supervised learning program 50 may be implemented, for example, by semiconductor integrated circuits, more specifically by ASICs (Application Specific Integrated Circuits), FPGAs (Field-Programmable Gate Arrays), etc.

[0041] Next, the operation of the self-supervised learning device 10 according to the first embodiment will be described. When the first molecular data is input to the self-supervised learning device 10 and the execution of self-supervised learning of NN24 is instructed, the self-supervised learning process shown in Figure 11 is executed in the self-supervised learning device 10. Note that the self-supervised learning process is an example of the self-supervised learning method of the disclosed technology.

[0042] In step S10, the generation unit 12 acquires the first molecular data input to the self-supervised learning device 10 and creates copies, one for generating the second molecular data and the other for generating the correct answer data. Next, in step S12, the generation unit 12 generates the second molecular data by shuffling the atoms contained in the first molecule in a complete permutation.

[0043] Next, in step S14, the acquisition unit 14 inputs the generated second molecular data into the NN24 to obtain prediction results regarding the molecular structure. Next, in step S16, the update unit 16 generates ground truth data based on the first molecular data. Next, in step S18, the update unit 16 backpropagates the difference between the prediction results and the ground truth data to the NN24 to update the parameters of the NN24.

[0044] Next, in step S20, the update unit 16 determines whether the termination condition for parameter updating has been met. If the termination condition is met, the process proceeds to step S22; otherwise, it returns to step S10. In step S22, the update unit 16 stores the NN24 with the final parameters set in a predetermined memory area, and the self-supervised learning process ends.

[0045] As described above, the self-supervised learning device according to the first embodiment generates second molecular data by substituting each of the multiple atoms contained in the first molecule with any of the atoms contained in the first molecule. The self-supervised learning device also inputs the second molecular data into a neural network (NN), which is a machine learning model that makes predictions about molecular structure, to obtain prediction results, and updates the parameters of the NN based on the comparison result between the correct data corresponding to the first molecule and the prediction result. In this way, instead of substituting atoms using atom species randomly selected from a set of atom species, the atoms contained in the original molecular data are shuffled to generate the second molecular data. As a result, since there are no unique atoms in the second molecule, the problem becomes predicting the original atoms from molecular structure such as interatomic distances and the angles of the positions of other atoms relative to a given atom. In other words, it becomes a problem that requires knowledge of the entire molecular structure rather than information about the vicinity of a particular atom, which accelerates the learning of essential features of the molecular structure and enables the generation of a highly accurate machine learning model through self-supervised learning.

[0046] In the first embodiment, the case in which the atoms contained in the first molecule are shuffled in a complete permutation to generate the second molecule data was described, but the method is not limited to this. In the case of shuffling in a complete permutation as in the first embodiment, the composition ratio, which is the ratio of the number of atoms of each atomic species to the total number of atoms contained in the first molecule, does not change before and after shuffling. Therefore, the composition ratio of atomic species can serve as a hint for prediction, potentially making prediction easier. For example, the atoms contained in the first molecule may be shuffled while allowing duplication of the atoms to be substituted by generating random integers from 0 to N-1, allowing for duplication.

[0047] For example, in the example in Figure 12, the index "5" in the original atom list is selected twice in the substituted atom list. As a result, the atom species "γ," which appeared only once in the original atom list, appears twice in the shuffled atom list. By changing the composition ratio before and after shuffling in this way, the problem becomes more difficult to predict, and a more accurate machine learning model can be generated.

[0048] Furthermore, while the first embodiment described the case in which all atoms in the first molecule are shuffled, the model is not limited to this. For example, the ratio of the number of atoms substituted by other atomic species to the total number of atoms in the first molecule (hereinafter referred to as the "substitution rate") may be set as a hyperparameter in advance, and only some of the atoms selected based on the substitution rate may be shuffled. However, when all atoms are shuffled as in the first embodiment, hyperparameters such as the substitution rate are unnecessary. Therefore, there is no need to tune hyperparameters, and a pre-trained model with a certain level of accuracy can be generated for any machine learning model or dataset.

[0049] <Second Embodiment> Next, a second embodiment will be described. In the self-supervised learning device according to the second embodiment, the same reference numerals are used for components similar to those in the self-supervised learning device 10 according to the first embodiment, and for functional parts with some common functions, the last two digits of the reference numerals are the same, and detailed explanations are omitted.

[0050] As shown in Figure 8, the self-supervised learning device 210 according to the second embodiment functionally includes a generation unit 212, an acquisition unit 14, and an update unit 16. Furthermore, NN24 is stored in a predetermined memory area of ​​the self-supervised learning device 210.

[0051] Here, atoms with a small proportion have a high probability of being replaced by other atoms after shuffling, while atoms with a large proportion have a low probability of being replaced by other atoms after shuffling. For example, in the original list of atoms shown in Figure 9A, the proportion of atom species α is 3 / 6=1 / 2, the proportion of atom species β is 2 / 6=1 / 3, and the proportion of atom species γ is 1 / 6. The probability that atom species α is replaced by another atom species is 3 / 6=1 / 2, the probability that atom species β is replaced by another atom species is 4 / 6=2 / 3, and the probability that atom species γ is replaced by another atom species is 5 / 6.

[0052] Thus, the fact that the probability of substitution by other atomic species differs depending on the composition ratio can serve as a hint during self-supervised learning and may facilitate prediction. Therefore, in the second embodiment, shuffling is performed in a way that does not cause variation in the probability of substitution by other atomic species for each atomic species.

[0053] The generation unit 212 selects atoms to be substituted so that the ratio of atoms to be substituted with different atomic species relative to the number of atoms of each atomic species contained in the first molecule (hereinafter referred to as the "substitution rate for each atomic species") satisfies predetermined conditions. The generation unit 212 then generates second molecular data by substituting the atomic species of the atoms to be substituted with other atomic species. The predetermined conditions may be, for example, that the substitution rate for each atomic species is a predetermined ratio for each composition ratio, or that it is a ratio randomly generated from a Gaussian distribution.

[0054] For example, suppose the generating unit 212 selects atoms corresponding to the elements shown by the thick lines in the original atom list shown in Figure 13A as the atoms to be replaced. In this case, the generating unit 212 may select atomic species β or γ in a 2:1 ratio for atomic species α, atomic species α or γ in a 3:1 ratio for atomic species β, and atomic species α or β in a 3:2 ratio for atomic species γ as the replacement atomic species.

[0055] Furthermore, the generation unit 212 may generate second molecular data by shuffling the atoms to be substituted in a complete permutation, similar to the generation unit 12 of the first embodiment. Figure 13B shows an example of the shuffled atom list in this case. Alternatively, the generation unit 212 may generate second molecular data by shuffling the atoms to be substituted, allowing for duplication. Figure 13C shows an example of the shuffled atom list in this case.

[0056] The self-supervised learning device 210 may be implemented, for example, by the computer 40 shown in Figure 10. The storage device 44 of the computer 40 stores a self-supervised learning program 250 that causes the computer 40 to function as the self-supervised learning device 210. The self-supervised learning program 250 includes a generation process control instruction 252, an acquisition process control instruction 54, and an update process control instruction 56. The storage device 44 also has an information storage area 60 in which information constituting the NN24 is stored.

[0057] The CPU 41 reads the self-supervised learning program 250 from the storage device 44, loads it into memory 43, and sequentially executes the control instructions contained in the self-supervised learning program 250. By executing the generation process control instruction 252, the CPU 41 operates as the generation unit 212 shown in Figure 8. The other control instructions are the same as those for the self-supervised learning program 50 in the first embodiment. As a result, the computer 40 that has executed the self-supervised learning program 250 functions as a self-supervised learning device 210.

[0058] Furthermore, the functions realized by the self-supervised learning program 250 may be implemented, for example, using semiconductor integrated circuits, or more specifically, ASICs, FPGAs, etc.

[0059] Next, the operation of the self-supervised learning device 210 according to the second embodiment will be described. When the first molecular data is input to the self-supervised learning device 210 and the execution of self-supervised learning of NN24 is instructed, the self-supervised learning process shown in Figure 14 is executed in the self-supervised learning device 210. In the self-supervised learning process in the second embodiment, the same process as the self-supervised learning process in the first embodiment (Figure 11) is given the same step number and a detailed explanation is omitted.

[0060] After step S10, in the next step S210, the generation unit 212 selects atoms to be substituted so that the substitution rate for each atom species satisfies predetermined conditions. Next, in step S212, the generation unit 212 generates second molecular data by substituting the atom species of the atoms to be substituted with other atom species. Thereafter, steps S14 to S22 are executed in the same manner as in the first embodiment, and the self-supervised learning process is completed.

[0061] As described above, the self-supervised learning device according to the second embodiment selects atoms to be substituted when generating second molecular data, such that the ratio of atoms to be substituted with different atomic species to the number of atoms of each atomic species contained in the first molecule satisfies a predetermined condition. The self-supervised learning device then generates second molecular data by substituting the atomic species of the atoms to be substituted with other atomic species. This makes the problem difficult to predict, as the substitution rate for each atomic species according to the composition ratio does not serve as a hint during self-supervised learning, and thus enables the generation of a highly accurate machine learning model through self-supervised learning.

[0062] <Third Embodiment> Next, a third embodiment will be described. In the self-supervised learning device according to the third embodiment, the same reference numerals are used for components similar to those in the self-supervised learning device 10 according to the first embodiment, and for functional parts that share some functions, the last two digits of the reference numerals are shared, and detailed explanations are omitted.

[0063] As shown in Figure 8, the self-supervised learning device 310 according to the third embodiment functionally includes a generation unit 312, an acquisition unit 14, and an update unit 16. Furthermore, NN24 is stored in a predetermined memory area of ​​the self-supervised learning device 310.

[0064] The generation unit 312 includes adjusting the substitution rate, which is the ratio of the number of atoms substituted with other atomic species to the number of atoms contained in the first molecule, so that it falls within a predetermined range. For example, when the generation unit 312 generates second molecule data by shuffling the atom list, as in the first embodiment, it compares the original atom list with the shuffled atom list and calculates the substitution rate. If the calculated substitution rate is above a predetermined upper threshold, i.e., if it has been shuffled too much, the generation unit 312 restores some of the atoms whose atomic species have changed back to their original atomic species. Also, if the calculated substitution rate is below a predetermined lower threshold, i.e., if it has not been shuffled enough, the generation unit 312 replaces some of the atoms whose atomic species have not changed with other atomic species.

[0065] The self-supervised learning device 310 may be implemented, for example, by the computer 40 shown in Figure 10. The storage device 44 of the computer 40 stores a self-supervised learning program 350 that causes the computer 40 to function as the self-supervised learning device 310. The self-supervised learning program 350 includes a generation process control instruction 352, an acquisition process control instruction 54, and an update process control instruction 56. The storage device 44 also has an information storage area 60 in which information constituting the NN24 is stored.

[0066] The CPU 41 reads the self-supervised learning program 350 from the storage device 44, loads it into memory 43, and sequentially executes the control instructions contained in the self-supervised learning program 350. By executing the generation process control instruction 352, the CPU 41 operates as the generation unit 312 shown in Figure 8. The other control instructions are the same as those for the self-supervised learning program 50 in the first embodiment. As a result, the computer 40 that has executed the self-supervised learning program 350 functions as a self-supervised learning device 310.

[0067] Furthermore, the functions realized by the self-supervised learning program 350 may be implemented, for example, by semiconductor integrated circuits, or more specifically, ASICs, FPGAs, etc.

[0068] Next, the operation of the self-supervised learning device 310 according to the third embodiment will be described. When the first molecular data is input to the self-supervised learning device 310 and the execution of self-supervised learning of NN24 is instructed, the self-supervised learning process shown in Figure 15 is executed in the self-supervised learning device 310. In the self-supervised learning process in the third embodiment, the same process as the self-supervised learning process in the first embodiment (Figure 11) is given the same step number and a detailed explanation is omitted.

[0069] After going through steps S10 to S12, in the next step S310, the generating unit 312 calculates the substitution rate, which is the ratio of the number of atoms substituted with other atomic species to the number of atoms contained in the first molecule. The generating unit 312 then determines whether the calculated substitution rate is within a predetermined range. If the substitution rate is within the predetermined range, the process proceeds to step S14. If the substitution rate is above the upper threshold or below the lower threshold of the predetermined range, the process proceeds to step S312.

[0070] In step S312, if the substitution rate is above the upper threshold, the generation unit 312 restores some of the atoms whose atomic species have changed back to their original atomic species. If the substitution rate is below the lower threshold, the generation unit 312 replaces some of the atoms whose atomic species have not changed with other atomic species. In this way, the generation unit 312 adjusts the substitution. Thereafter, steps S14 to S22 are executed as in the first embodiment, and the self-supervised learning process is completed.

[0071] As described above, the self-supervised learning device according to the third embodiment adjusts the substitution rate, which is the ratio of the number of atoms substituted with other atomic species to the number of atoms contained in the first molecule, to a predetermined range when generating second molecular data. This allows for adjustment of the degree of unpredictability during self-supervised learning and enables the generation of an appropriate machine learning model.

[0072] In the third embodiment, the case in which the first molecular data is shuffled to generate the second molecular data and then the substitution adjustments are made was described, but the invention is not limited to this. Before shuffling, atoms to be substituted may be selected so that the substitution rate falls within a predetermined range, and the atomic species of the selected atoms may be substituted with other atomic species.

[0073] <Fourth Embodiment> Next, a fourth embodiment will be described. In the self-supervised learning device according to the fourth embodiment, the same reference numerals are used for components similar to those in the self-supervised learning device 10 according to the first embodiment, and for functional parts that share some functions, the last two digits of the reference numerals are shared, and detailed explanations are omitted.

[0074] As shown in Figure 8, the self-supervised learning device 410 according to the fourth embodiment functionally includes a generation unit 412, an acquisition unit 14, and an update unit 16. Furthermore, NN24 is stored in a predetermined memory area of ​​the self-supervised learning device 410.

[0075] The generation unit 412 generates multiple candidate second molecular data by repeatedly substituting each of the multiple atoms contained in the first molecule with any of the atoms contained in the first molecule. Then, the generation unit 412 selects the best candidate as the second molecular data based on an evaluation according to the substitution rate of each atom species. The generation unit 412 may perform an evaluation according to the substitution rate of each atom species based on how well the substitution rate of each atom species satisfies predetermined conditions. For example, the generation unit 412 may evaluate candidates as better the less variation there is in the substitution rate of each atom species, or it may evaluate candidates as better the more the substitution rate of each atom species follows a Gaussian distribution.

[0076] The self-supervised learning device 410 may be implemented, for example, by the computer 40 shown in Figure 10. The storage device 44 of the computer 40 stores a self-supervised learning program 450 that causes the computer 40 to function as the self-supervised learning device 410. The self-supervised learning program 450 includes a generation process control instruction 452, an acquisition process control instruction 54, and an update process control instruction 56. The storage device 44 also has an information storage area 60 in which information constituting the NN24 is stored.

[0077] The CPU 41 reads the self-supervised learning program 450 from the storage device 44, loads it into memory 43, and sequentially executes the control instructions contained in the self-supervised learning program 450. By executing the generation process control instruction 452, the CPU 41 operates as the generation unit 412 shown in Figure 8. The other control instructions are the same as those for the self-supervised learning program 50 in the first embodiment. As a result, the computer 40 that has executed the self-supervised learning program 450 functions as a self-supervised learning device 410.

[0078] Furthermore, the functions realized by the self-supervised learning program 450 may be implemented, for example, by semiconductor integrated circuits, or more specifically, ASICs, FPGAs, etc.

[0079] Next, the operation of the self-supervised learning device 410 according to the fourth embodiment will be described. When the first molecular data is input to the self-supervised learning device 410 and the execution of self-supervised learning of NN24 is instructed, the self-supervised learning process shown in Figure 16 is executed in the self-supervised learning device 410. In the self-supervised learning process in the fourth embodiment, the same process as the self-supervised learning process in the first embodiment (Figure 11) is given the same step number and a detailed explanation is omitted.

[0080] After step S10, in the next step S410, the generation unit 412 performs a process multiple times in which it replaces each of the multiple atoms contained in the first molecule with one of the atoms contained in the first molecule, thereby generating multiple candidate second molecule data. Next, in step S412, the generation unit 412 selects the best candidate as the second molecule data based on an evaluation according to the substitution rate for each atom species. Thereafter, steps S14 to S22 are executed in the same manner as in the first embodiment, and the self-supervised learning process is completed.

[0081] As described above, the self-supervised learning apparatus according to the fourth embodiment generates a plurality of candidate second molecular data sets and selects the best candidate as the second molecular data set based on an evaluation according to the substitution rate for each atomic species. This makes it possible to generate second molecular data suitable for self-supervised learning and to generate a machine learning model with high accuracy.

[0082] Here, we will explain the verification results of the effects of each of the above embodiments. In this verification, the following methods were tested.

[0083] Comparison Method 1: A method for supervised learning of a neural network using only labeled data (without self-supervised learning) Comparison Method 2: A method that fine-tunes a neural network (NN) pre-trained using self-supervised learning (predicting whether or not atoms have been substituted) from Reference Method 1, using labeled data. Method 1: A method for fine-tuning a neural network (NN) pre-trained (predicting the original atoms) using the method of the first embodiment with labeled data.

[0084] Furthermore, in comparison method 2, after confirming that it is more accurate than the task of predicting the original atoms, the task of predicting whether or not atoms have been substituted is adopted.

[0085] In this validation, assuming that a small amount of data is used for fine-tuning, evaluation was performed using a portion of a publicly available dataset. For self-supervised learning, 460,328 training data points and 24,943 validation data points were used. For fine-tuning, 1,120 training data points and 2,624 validation data points were extracted from the total of 45,890 data points. Furthermore, a pre-trained model was generated by performing self-supervised learning for 4 epochs, and a machine learning model that predicts energy from molecular structure was generated by performing fine-tuning for 200 epochs. Note that in comparison method 1, only the machine learning equivalent to fine-tuning was performed.

[0086] The mean absolute error between the energy predicted for the validation data by the machine learning model generated through fine tuning and the energy label was used as an evaluation value for comparison. In other words, a lower evaluation value indicates higher accuracy. The table below shows the evaluation values ​​after fine tuning. As shown in the table below, an improvement in accuracy was observed with this method.

[0087] [Table 1]

[0088] In each of the above embodiments, the self-supervised learning program is pre-stored (installed) in the storage device, but this is not limited to this. The program relating to the disclosed technology may be provided in a form stored on a storage medium such as a CD-ROM, DVD-ROM, or USB memory.

[0089] The following additional information is disclosed regarding each of the above embodiments.

[0090] (Note 1) Data is generated showing a second molecule in which each of the multiple atoms contained in the first molecule is replaced with any of the atoms contained in the first molecule. The data representing the second molecule is input into a machine learning model that makes predictions about the molecular structure to obtain prediction results. Based on the comparison result between the ground truth data corresponding to the first molecule and the prediction result, the parameters of the machine learning model are updated. A self-supervised learning program that causes a computer to perform a process that includes the following.

[0091] (Note 2) The process for generating data representing the second molecule is a self-supervised learning program as described in Appendix 1, which includes shuffling the atoms contained in the first molecule in a complete permutation.

[0092] (Note 3) The process for generating data representing the second molecule is a self-supervised learning program as described in Appendix 1, which includes shuffling the atoms contained in the first molecule while allowing duplication of the atoms to be substituted.

[0093] (Note 4) The process for generating data representing the second molecule is a self-supervised learning program as described in Appendix 1, which includes selecting atoms to be substituted such that the ratio of the number of atoms to be substituted with a different atomic species to the number of atoms of each atomic species contained in the first molecule satisfies a predetermined condition, and substituting the atomic species of the atoms to be substituted with other atomic species.

[0094] (Note 5) The process for generating data representing the second molecule is a self-supervised learning program according to any one of the appendices 1 to 4, which includes adjusting the ratio of the number of atoms substituted with other atomic species to the number of atoms contained in the first molecule so that it falls within a predetermined range.

[0095] (Note 6) The process for generating data representing the second molecule involves performing a process multiple times in which each of the multiple atoms contained in the first molecule is replaced with one of the atoms contained in the first molecule, thereby generating multiple candidate data representing the second molecule, and selecting the best candidate as the data representing the second molecule based on an evaluation according to the ratio of the number of replaced atoms for each atom species, as described in any one of the appendices 1 to 5 of the self-supervised learning program.

[0096] (Note 7) The machine learning model is a self-supervised learning program as described in any one of the appendices 1 to 6, which predicts whether each atom in the second molecule is substituted for each atom in the first molecule.

[0097] (Note 8) The machine learning model is a self-supervised learning program as described in any one of the appendices 1 to 6, which predicts the original atoms of the atoms in the second molecule that have been substituted from the atoms in the first molecule.

[0098] (Note 9) Data is generated showing a second molecule in which each of the multiple atoms contained in the first molecule is replaced with any of the atoms contained in the first molecule. The data representing the second molecule is input into a machine learning model that makes predictions about the molecular structure to obtain prediction results. Based on the comparison result between the ground truth data corresponding to the first molecule and the prediction result, the parameters of the machine learning model are updated. A self-supervised learning method in which a computer performs processes that include the following.

[0099] (Note 10) The process for generating data representing the second molecule is a self-supervised learning method as described in Appendix 9, which includes shuffling the atoms contained in the first molecule in a complete permutation.

[0100] (Note 11) The process for generating data representing the second molecule is a self-supervised learning method as described in Appendix 9, which includes shuffling the atoms contained in the first molecule while allowing duplication of the atoms to be substituted.

[0101] (Note 12) The process for generating data representing the second molecule includes selecting atoms to be substituted such that the ratio of the number of atoms to be substituted with a different atomic species to the number of atoms of each atomic species contained in the first molecule satisfies a predetermined condition, and substituting the atomic species of the atoms to be substituted with other atomic species, as described in Appendix 9.

[0102] (Note 13) The self-supervised learning method described in any one of the appendices 9 to 12, wherein the process for generating data representing the second molecule includes adjusting the ratio of the number of atoms substituted with other atomic species to the number of atoms contained in the first molecule so that it falls within a predetermined range.

[0103] (Note 14) The process for generating data representing the second molecule involves repeatedly performing a process in which each of the multiple atoms contained in the first molecule is replaced with any of the atoms contained in the first molecule, thereby generating multiple candidate data representing the second molecule, and selecting the best candidate as the data representing the second molecule based on an evaluation according to the ratio of the number of replaced atoms for each atom species, as described in any one of the appendices 9 to 13.

[0104] (Note 15) The machine learning model predicts whether each atom in the second molecule is substituted for each atom in the first molecule, according to the self-supervised learning method described in any one of the appendices 9 to 14.

[0105] (Note 16) The machine learning model predicts the original atoms of the atoms in the second molecule that have been replaced by atoms in the first molecule, according to the self-supervised learning method described in any one of the appendices 9 to 14.

[0106] (Note 17) A generation unit that generates data indicating a second molecule in which each of the multiple atoms contained in the first molecule is replaced with any of the atoms contained in the first molecule, An acquisition unit that inputs data representing the second molecule into a machine learning model that makes predictions about the molecular structure and obtains prediction results, An update unit updates the parameters of the machine learning model based on a comparison result between the ground truth data corresponding to the first molecule and the prediction result, A self-supervising learning device that includes [this].

[0107] (Note 18) The generation unit is a self-supervised learning device as described in Appendix 17, which shuffles the atoms contained in the first molecule in a complete permutation.

[0108] (Note 19) The generation unit is a self-supervised learning device as described in Appendix 17, which shuffles the atoms contained in the first molecule, allowing for duplication of the atoms to be substituted.

[0109] (Note 20) The generating unit selects atoms to be substituted such that the ratio of atoms to be substituted with different atomic species to the number of atoms of each atomic species contained in the first molecule satisfies a predetermined condition, and replaces the atomic species of the atoms to be substituted with other atomic species, as described in Appendix 17. [Explanation of Symbols]

[0110] 10, 210, 310, 410 Self-supervised learning devices 12, 212, 312, 412 generation section 14 Acquisition Department 16 Update section 22 Atomic Library 24 NN 40 Computers 41 CPU 42 GPU 43 memory 44 Storage device 45 Input / Output Devices 46 R / W device 47 Communication I / F 48 bus 49 Storage medium 50, 250, 350, 450 Self-supervised learning programs 52, 252, 352, 452 Generation process control instructions 54 Acquisition process control instruction 56 Update process control instructions 60 Information storage area

Claims

1. Data is generated showing a second molecule in which each of the multiple atoms contained in the first molecule is replaced with any of the atoms contained in the first molecule. The data representing the second molecule is input into a machine learning model that makes predictions about the molecular structure to obtain prediction results. Based on the comparison result between the ground truth data corresponding to the first molecule and the prediction result, the parameters of the machine learning model are updated. A self-supervised learning program that causes a computer to perform a process that includes the following.

2. The self-supervised learning program according to claim 1, wherein the process for generating data representing the second molecule includes shuffling the atoms contained in the first molecule in a complete permutation.

3. The self-supervised learning program according to claim 1, wherein the process for generating data representing the second molecule includes shuffling the atoms contained in the first molecule, allowing for duplication of the atoms to be substituted.

4. The self-supervised learning program according to claim 1, wherein the process for generating data representing the second molecule includes selecting atoms to be substituted such that the ratio of the number of atoms to be substituted with a different atomic species to the number of atoms of each atomic species contained in the first molecule satisfies a predetermined condition, and substituting the atomic species of the atoms to be substituted with other atomic species.

5. A self-supervised learning program according to any one of claims 1 to 4, wherein the process for generating data representing the second molecule includes adjusting the ratio of the number of atoms substituted with other atomic species to the number of atoms contained in the first molecule so that it falls within a predetermined range.

6. A self-supervised learning program according to any one of claims 1 to 4, wherein the process for generating data representing the second molecule involves performing a process multiple times in which each of the multiple atoms contained in the first molecule is replaced with any of the atoms contained in the first molecule, thereby generating multiple candidate data representing the second molecule, and selecting the best candidate as data representing the second molecule based on an evaluation according to the ratio of the number of replaced atoms for each atom species.

7. The self-supervised learning program according to any one of claims 1 to 4, wherein the machine learning model predicts whether each of the atoms contained in the second molecule is substituted for each of the atoms contained in the first molecule.

8. The self-supervised learning program according to any one of claims 1 to 4, wherein the machine learning model predicts the original atom of the atom that was replaced from the atom in the first molecule among the atoms in the second molecule.

9. Data is generated showing a second molecule in which each of the multiple atoms contained in the first molecule is replaced with any of the atoms contained in the first molecule. The data representing the second molecule is input into a machine learning model that makes predictions about the molecular structure to obtain prediction results. Based on the comparison result between the ground truth data corresponding to the first molecule and the prediction result, the parameters of the machine learning model are updated. A self-supervised learning method in which a computer performs processes that include the following.

10. A generation unit that generates data indicating a second molecule in which each of the multiple atoms contained in the first molecule is replaced with any of the atoms contained in the first molecule, An acquisition unit that inputs the data representing the second molecule into a machine learning model that makes predictions about the molecular structure and obtains prediction results, An update unit updates the parameters of the machine learning model based on a comparison result between the correct data corresponding to the first molecule and the prediction result, A self-supervising learning device that includes [this].

Images