PREDICTION METHOD WITH LIMITED INPUT VARIABLES USING ARTIFICIAL NEURAL NETWORK

Abstract

To disclose a method for performing prediction using an artificial neural network.SOLUTION: Specifically, according to the present disclosure, a computing device obtains input data comprising the second number of input variables and generates a prediction value by inputting the input data to a graph neural network model. At this time, the graph neural network model corresponds to an artificial neural network model learned on the basis of the relationship among the first number of input variables being greater than or equal to the second number.SELECTED DRAWING: Figure 3

Description

[Technical field]

[0001] The present disclosure relates to a prediction method using an artificial neural network in an environment with limited input variables, and more particularly to a method for generating a prediction value for input data using a graph neural network model trained based on relationships between input variables that are equal to or greater than the number of input variables contained in the input data. [Background technology]

[0002] With the development of artificial intelligence technology, artificial neural network models are being used to perform various tasks in many fields.

[0003] Among them, the graph neural network is a type of artificial neural network that receives data including graphs as input and generates predicted values, and is an artificial neural network specialized in generating predicted values ​​by reflecting the relationship between input variables. A characteristic of such a graph neural network is that the number of input variables that make up the training data must be the same as the number of input variables that make up the input data in the inference stage.

[0004] However, when performing a task using an actual graph neural network model, there are various input variables available in the learning process, but the input variables input in the inference process may be limited. For example, in the case of a graph neural network that receives input of a patient's medical data and generates diagnostic information for the patient, medical data including various input variables measured from a patient in an intensive care unit can be acquired and learned in the learning process, but when attempting to generate diagnostic information for a patient in a general ward in the inference process, the number of input variables of the input medical data may be less than the number of input variables of the medical data acquired from the intensive care unit.

[0005] In order to solve this problem, a method has been used in which only some input variables that can be commonly obtained in all situations are selected from among possible input variables, and then the input variables of the training data are limited to the selected input variables, and then a graph neural network model is trained using only the selected input variables. However, this method cannot reflect relationship information between input variables included in the training data and other input variables not included in the training data, and as a result, the characteristic of the graph neural network that reflects relationship information between data cannot be utilized, resulting in an inaccurate inference result.

[0006] Therefore, there is a need in the industry for a method that can generate accurate predictions for input data in an environment with limited input variables.

[0007] Korean Patent No. 2485944 discloses a graph encoding method in a transformer neural network. Summary of the Invention [Problem to be solved by the invention]

[0008] The present disclosure has been made in response to the above-mentioned background art, and aims to generate a prediction value for input data using a graph neural network model trained based on relationships between input variables that are equal to or greater than the number of input variables in the input data.

[0009] On the other hand, the technical problem that the present disclosure aims to achieve is not limited to the technical problem described above, and may include various technical problems within the scope that is obvious to a person of ordinary skill in the art from the content described below. [Means for solving the problem]

[0010] According to an embodiment of the present disclosure for achieving the above-mentioned object, a method for performing prediction using an artificial neural network via a computing device is disclosed, the method including the steps of obtaining input data including a second number of input variables, and generating a prediction based on inputting the input data to a graph neural network model, the graph neural network model may be an artificial neural network model trained based on relationships between a first number of input variables, the first number being greater than or equal to the second number.

[0011] In one embodiment, the graph neural network model may be an artificial neural network model trained using training data including the first number of input variables.

[0012] In one embodiment, the graph neural network model may be an artificial neural network model trained by generating a first adjacency matrix based on the first number of input variables, and training the graph neural network model based on the training data and the first adjacency matrix.

[0013] In one embodiment, the step of generating a predicted value based on inputting data for the second number of input variables into a graph neural network model may include: generating a second adjacency matrix based on the second number of input variables; and generating a predicted value for the input data based on the input data, the second adjacency matrix, and learned parameters of the graph neural network model.

[0014] In one embodiment, the step of generating a second adjacency matrix based on the input variables of the input data may include: generating a first adjacency matrix based on relationships between the first number of input variables; and masking elements of the first adjacency matrix that do not represent relationships between a second number of input variables.

[0015] In one embodiment, generating a first adjacency matrix based on relationships between the first number of input variables may include: generating a first adjacency matrix based on relationships between the first number of input variables and relative importance between the first number of input variables.

[0016] In one embodiment, the adjacency matrix may be a matrix representing a graph composed of nodes and edges.

[0017] In one embodiment, the graph may be a graph generated including the steps of determining nodes based on input variables, and determining edges connecting each node based on relationships between the input variables.

[0018] In one embodiment, the graph neural network model may include an artificial neural network model that receives input of a patient's time-series medical data and outputs a diagnosis result.

[0019] According to an embodiment of the present disclosure for achieving the above-mentioned object, a computer program is disclosed for performing operations for performing prediction using an artificial neural network via a computing device, the operations including: obtaining input data including a second number of input variables, and generating a prediction based on inputting the input data to a graph neural network model, where the graph neural network model may be an artificial neural network model trained based on relationships between a first number of input variables, the first number being greater than or equal to the second number.

[0020] According to an embodiment of the present disclosure for achieving the above-mentioned object, a computing device for performing prediction using an artificial neural network is disclosed. The computing device includes one or more processors and a memory, and the processor is capable of obtaining input data including a second number of input variables and generating a prediction value based on inputting the input data into a graph neural network model. The graph neural network model may be an artificial neural network model trained based on relationships between a first number of input variables, the first number being greater than or equal to the second number. Effect of the Invention

[0021] The present disclosure has an effect of making predictions by an artificial neural network model more accurate even when data with limited input variables is input. In particular, the present disclosure has an effect of generating a prediction value for input data including a second number of input variables, which is less than the first number, after a graph neural network model is trained based on the relationships between a first number of input variables, thereby generating an accurate prediction value that further takes into account the relationships between variables not included in the input data. [Brief description of the drawings]

[0022] [Figure 1] FIG. 1 is a block diagram of a computing device for performing predictions using an artificial neural network according to one embodiment of the present disclosure. [Diagram 2] FIG. 2 is a schematic diagram illustrating a network function according to an embodiment of the present disclosure. [Diagram 3] 1 is a flowchart illustrating a process for training a graph neural network model according to one embodiment of the present disclosure. [Figure 4] FIG. 1 is a conceptual diagram illustrating training data for a graph neural network model according to an embodiment of the present disclosure. [Diagram 5]FIG. 1 is a conceptual diagram illustrating input data for a graph neural network model according to an embodiment of the present disclosure. [Figure 6] FIG. 1 illustrates a simplified general schematic diagram of an exemplary computing environment in which embodiments of the present disclosure can be implemented. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0023] The present disclosure discloses a method for utilizing a graph neural network model trained based on relationships between input variables that are greater than or equal to the number of input variables in the input data, and generating a prediction based on inputting the input data into the graph neural network model.

[0024] Various embodiments are described below with reference to the drawings, in which like reference numerals are used to represent like components throughout the drawings. Various descriptions are provided herein to facilitate understanding of the present disclosure. However, the embodiments can undoubtedly be practiced without these specific descriptions.

[0025] As used herein, terms such as "component," "module," and "system" refer to computer-related entities, hardware, firmware, software, a combination of software and hardware, or software execution. For example, a component can be, but is not limited to, a procedure running on a processor, a processor, an object, a thread of execution, a program, and / or a computer. For example, both an application running on a computing device and the computing device can be a component. One or more components can reside within a processor and / or thread of execution, and a component can be localized within one computer or distributed across two or more computers. Such components can also execute from various computer-readable media having various data structures stored therein. Components can communicate through local and / or remote processing, etc., by signals (e.g., data from one component interacting with other components in a local system, a distributed system, and / or data transmitted over a network such as the Internet to other systems), for example, with one or more data packets.

[0026] The term "or" is intended to mean an inclusive "or" rather than an exclusive "or." That is, unless otherwise specified or clear from the context, "X utilizes A or B" is intended to mean one of the natural inclusive permutations. That is, if X utilizes A; X utilizes B; or X utilizes both A and B, then "X utilizes A or B" can apply to any of these. Additionally, the term "and / or" as used herein should be understood to refer to and include all possible combinations of one or more of the associated listed items.

[0027] Additionally, the predicate "comprises" and / or the modifier "comprises" should be understood to mean that the feature and / or component is present. However, the predicate "comprises" and / or the modifier "comprises" should be understood to not exclude the presence or addition of one or more other further features, components and / or groups thereof. Additionally, in the present specification and claims, the singular should generally be construed to mean "one or more," unless a specific number is specified or the context is clear to indicate the singular form.

[0028] Furthermore, the term "at least one of A or B" should be interpreted as meaning "a case in which only A is included," "a case in which only B is included," or "a case in which a combination of A and B is included."

[0029] Those skilled in the art should further appreciate that the various exemplary logical blocks, configurations, modules, circuits, means, logic, and algorithm steps described in accordance with the embodiments disclosed herein can be realized by electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the various exemplary components, blocks, configurations, means, logic, modules, circuits, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented as hardware or software depends on the particular application and design limitations of the overall system. Skilled artisans can implement the described functionality in various ways for each particular application. However, such implementation decisions should not be interpreted as departing from the scope of the present disclosure.

[0030] The description of the embodiments set forth herein is provided to enable one of ordinary skill in the art to use or practice the present disclosure. Various modifications to these embodiments will be apparent to those of ordinary skill in the art, and the general principles defined herein may be applied to other embodiments without departing from the scope of the present disclosure. Thus, the present disclosure is not intended to be limited by the embodiments set forth herein, but is to be accorded the broadest scope consistent with the principles and novel features disclosed herein.

[0031] In the present disclosure, a graph neural network model may refer to a type of artificial neural network model for processing graph-formatted data. The graph neural network model can extract node characteristics from a graph structure composed of nodes and edges, and perform tasks such as classification and prediction. The graph neural network model can generally learn a latent representation of a node using information on the node and its adjacent nodes. To this end, the graph neural network model can update the latent representation of the node while propagating information by stacking multiple layers. The graph neural network model is a model suitable for graph-formatted data, and has the characteristic of well reflecting the relationship between data. By using the graph neural network model, it can be used in various fields such as social network analysis, molecular structure analysis, and geographic information systems.

[0032] In the present disclosure, an adjacency matrix may be a matrix that expresses relationship information between multiple pieces of data. In general, when the number of pieces of data is N, the adjacency matrix may be a matrix of size N×N, and the elements constituting each matrix may represent relationship information between the data representing the row and the data representing the column of the element. In general, in an adjacency matrix, if two pieces of data are related, they can be expressed as 1, and if they are not related, they can be expressed as 0.

[0033] In the present disclosure, the input variables may refer to variables constituting data, particularly time-series data. For example, if the input data is a record of a patient's heart rate, respiration rate, systolic blood pressure, body temperature, and arterial blood oxygen saturation (SpO2) value over time, the variables of the input data are {heart rate, respiration rate, systolic blood pressure, body temperature, SpO2}, and the number of variables of the input data is 5.

[0034] FIG. 1 is a block diagram of a computing device for performing predictions using an artificial neural network according to one embodiment of the present disclosure.

[0035] 1 is merely a simplified example, and in one embodiment of the present disclosure, the computer device (100) may include other components for implementing the computing environment of the computer device (100), and the computer device (100) may be configured with only some of the disclosed components.

[0036] The computing device (100) may include a processor (110), a memory (130), and a network portion (150).

[0037] In one embodiment of the present disclosure, the processor (100) may be configured with one or more cores and may include a processor for data analysis and deep learning, such as a computing central processing unit (CPU), a general purpose graphics processing unit (GPGPU), and a tensor processing unit (TPU). The processor (110) may read a computer program stored in the memory (130) and execute data processing for machine learning in one embodiment of the present disclosure. According to one embodiment of the present disclosure, the processor (110) may perform calculations for learning a neural network. In deep learning (DL), the processor (110) may execute calculations for learning a neural network, such as processing input data for learning, extracting features from the input data, calculating errors, and updating the weights of the neural network using backpropagation.

[0038] At least one of the CPU, GPGPU, and TPU of the processor (110) can process the learning of the network function. For example, the CPU and the GPGPU can both perform the learning of the network function or the classification of data using the network function. In one embodiment of the present disclosure, the processors of multiple computing devices can be used together to perform the learning of the network function or the classification of data using the network function. In one embodiment of the present disclosure, the computer program executed in the computing device can be a program executable by the CPU, the GPGPU, or the TPU.

[0039] The processor 110 can obtain input data. The input data of the present disclosure may be data having one single value, or may be time-series data including data by time point for each input variable.

[0040] The processor 110 may generate a predicted value based on inputting the input data to a graph neural network model, which is a trained artificial neural network model. In this case, the graph neural network model may be an artificial neural network model trained based on training data including a first number of input variables. When the number of input variables included in the input data is a second number, the first number may be greater than or equal to the second number. For example, the graph neural network model of the present disclosure may be trained with training data including four input variables {A, B, C, D}, and the processor 110 may input input data including three input variables {A, B, C}, which are part of the input variables included in the training data, to the graph neural network model to generate a predicted value.

[0041] When the graph neural network model of the present disclosure is trained based on training data including a first number of input variables, the graph neural network model can learn relationship information between the first number of input variables and reflect the relationship information in generating a predicted value. For example, the graph neural network model can be trained based on training data including four input variables {A, B, C, D}, and the weights, i.e., learnable parameters, of the graph neural network model can be updated to reflect information regarding the relationship between A and B, the relationship between A and C, the relationship between A and D, the relationship between B and C, the relationship between B and D, and the relationship between C and D. In general, the more input variables in the data, the higher the accuracy of the predicted value generated by the graph neural network tends to be.

[0042] The processor 110 may train the graph neural network model using training data including the first number of input variables. A specific method for training the graph neural network model will be described later with reference to FIG.

[0043] The processor 110 may generate a first adjacency matrix that is an adjacency matrix associated with training data in the process of training the graph neural network model. Meanwhile, the processor 110 may generate a second adjacency matrix that is an adjacency matrix associated with input data in the process of generating a predicted value using the graph neural network model. A specific method of generating the first and second adjacency matrices will be described later with reference to FIGS. 4 and 5.

[0044] Conventionally, when the input data has few input variables, an artificial neural network model trained using only the input variables is used to generate a predicted value for the input data. For example, when the input data has three input variables {A, B, C}, even if the training data has four input variables {A, B, C, D}, only the three input variables {A, B, C} corresponding to the input variables of the input data are selected to train a graph neural network model. However, in the case of an artificial neural network model trained in this way, a problem occurs in that a predicted value is generated that does not reflect at all the relationship between the input variables of the input data and the input variables not included in the input data.

[0045] In the present disclosure, a graph neural network model is trained using relationship information between a large number or the same number of input variables (first number) based on training data including a large number or the same number of input variables, and the weights of the graph neural network model are updated to reflect the relationship information between the large number or the same number of input variables. Then, the weights reflecting the relationship information between the large number or the same number of input variables are used as they are to generate a predicted value for input data including a small number or the same number of input variables (second number). Therefore, there is an effect of generating more accurate predicted information than the conventional method using a model trained with the same number of input variables as the input data.

[0046] In one embodiment of the present disclosure, the memory (130) may include at least one type of storage medium among flash memory type, hard disk type, multimedia card micro type, card type memory (e.g., SD or XD memory, etc.), Random Access Memory (RAM), Static Random Access Memory (SRAM), Read-Only Memory (ROM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Programmable Read-Only Memory (PROM), magnetic memory, magnetic disk, and optical disk. The computing device (100) may also operate in conjunction with a web storage that performs the storage function of the memory (130) over the Internet. The above description of the memory is merely exemplary, and the present disclosure is not limited thereto.

[0047] In one embodiment of the present disclosure, the network unit (150) can use various wired communication systems such as a Public Switched Telephone Network (PSTN), x Digital Subscriber Line (xDSL), Rate Adaptive DSL (RADSL), Multi Rate DSL (MDSL), Very High Speed ​​DSL (VDSL), Universal Asymmetric DSL (UADSL), High Bit Rate DSL (HDSL), and a Local Area Network (LAN).

[0048] In addition, the network unit 150 in the present specification can use various wireless communication systems such as Code Division Multi Access (CDMA), Time Division Multi Access (TDMA), Frequency Division Multi Access (FDMA), Orthogonal Frequency Division Multi Access (OFDMA), Single Carrier-FDMA (SC-FDMA), and other systems. The network unit 150 in an embodiment of the present disclosure can operate in conjunction with any known wired or wireless communication system having any type.

[0049] The techniques described herein may be used in the networks mentioned above as well as other networks.

[0050] FIG. 2 is a schematic diagram illustrating network functions used to provide diagnostic relevant information for medical data in one embodiment of the present disclosure.

[0051] Throughout this specification, the terms computational model, neural network, network function, and neural network may be used interchangeably. A neural network is often composed of a collection of interconnected computational units, generally called nodes. Such nodes may also be referred to as neurons. A neural network is composed of at least one or more nodes. The nodes (or neurons) that make up a neural network may be interconnected by one or more links.

[0052] In a neural network, one or more nodes connected via links can form a relationship of input node and output node relative to each other. The concept of input node and output node is relative, and any node that is an output node for one node can be an input node in relation to another node, and vice versa. As mentioned above, the relationship between input node and output node can be established around the link. One or more output nodes can be connected to one input node via links, and vice versa.

[0053] In the relationship between an input node and an output node connected via a link, the data of the output node can be determined based on the data input to the input node. Here, the node interconnecting the input node and the output node can have a weight. The weight can be variable and can be changed by a user or an algorithm so that the neural network performs a desired function. For example, when one or more input nodes are interconnected to one output node by each link, the output node can determine the value of the output node based on the value input to the input node connected to the output node and the weight set to the link corresponding to each input node.

[0054] As mentioned above, in a neural network, one or more nodes are interconnected through one or more links to form a relationship between an input node and an output node in the neural network. In a neural network, the characteristics of the neural network can be determined by the number of nodes and links, the correlation between the nodes and links, and the weight value assigned to each link. For example, if there are two neural networks that have the same number of nodes and links but different link weight values, the two neural networks can be recognized as different.

[0055] A neural network may be composed of a set of one or more nodes. A subset of the nodes in a neural network may form a layer. A portion of the nodes in a neural network may form a layer based on their distance from a first input node. For example, a set of nodes whose distance from a first input node is n may form an n-layer. The distance from the first input node may be defined based on the minimum number of links that must be traversed to reach the node from the first input node. However, this definition of a layer is given arbitrarily for the purpose of explanation, and the configuration of layers in a neural network may be defined in a manner different from that described above. For example, a layer of nodes may be defined based on their distance from a final output node.

[0056] The first input node may refer to one or more nodes in a neural network to which data is directly input without passing through a link in relation to other nodes. Alternatively, the first input node may refer to a node that does not have other input nodes connected via a link in relation between nodes based on links in a neural network. Similarly, the final output node may refer to one or more nodes in a neural network that do not have an output node in relation to other nodes. In addition, the hidden node may refer to a node that is neither a first input node nor a final output node and that constitutes a neural network.

[0057] The neural network according to an embodiment of the present disclosure may be a neural network in which the number of nodes in the input layer is the same as the number of nodes in the output layer, and the number of nodes decreases once and then increases again as the input layer progresses to the hidden layer. The neural network according to an embodiment of the present disclosure may be a neural network in which the number of nodes in the input layer is smaller than the number of nodes in the output layer, and the number of nodes decreases as the input layer progresses to the hidden layer. Furthermore, the neural network according to another embodiment of the present disclosure may be a neural network in which the number of nodes in the input layer is greater than the number of nodes in the output layer, and the number of nodes increases as the input layer progresses to the hidden layer. The neural network according to another embodiment of the present disclosure may be a neural network in which the above-mentioned neural networks are combined.

[0058] Deep neural networks (DNNs) can refer to neural networks that include multiple hidden layers in addition to input and output layers. Deep neural networks can be used to understand the latent structures of data. In other words, they can understand the latent structures of photos, text, video, audio, and music (e.g., whether an object is in the photo, what is the content and emotion of the text, what is the content and emotion of the audio, etc.). Deep neural networks can include convolutional neural networks (CNNs), recurrent neural networks (RNNs), autoencoders, generative adversarial networks (GANs), restricted boltzmann machines (RBMs), deep belief networks (DBNs), Q networks, U networks, Siamese networks, generative adversarial networks (GANs), etc. The above-mentioned deep neural networks are merely examples and the present disclosure is not limited thereto.

[0059] In one embodiment of the present disclosure, the network function may include an autoencoder. The autoencoder may be a type of artificial neural network for outputting output data similar to input data. The autoencoder may include at least one hidden layer, and an odd number of hidden layers may be disposed between the input and output layers. The number of nodes in each layer may decrease from the number of nodes in the input layer toward an intermediate layer called a bottleneck layer (encode), and may be expanded from the bottleneck layer toward the output layer (symmetrical to the input layer) in a manner symmetrical to the reduction. The autoencoder may perform nonlinear dimensionality reduction. The number of input layers and output layers may correspond to the dimensions after preprocessing of the input data. In the autoencoder structure, the number of nodes in the hidden layer included in the encoder may decrease as it gets farther from the input data. If the number of nodes in the bottleneck layer (the layer with the fewest number of nodes located between the encoder and the decoder) is too small, there is a possibility that a sufficient amount of information may not be transmitted, so it may be maintained at a certain number or more (e.g., more than half of the input layer, etc.).

[0060] Neural networks can be trained by at least one of supervised learning, unsupervised learning, semi-supervised learning, or reinforcement learning. Training of a neural network can be a process of providing the neural network with knowledge to perform a specific operation. A neural network can be trained in a direction that minimizes the error of the output. In training of a neural network, a process is performed in which training data is repeatedly input to the neural network, the error of the neural network output and the target regarding the training data is calculated, and the weight value of each node of the neural network is updated by backpropagating the error of the neural network from the output layer of the neural network to the input layer as a direction to reduce the error. In the case of supervised learning, training data in which the correct answer is labeled is used for each training data (i.e., labeled training data), and in the case of unsupervised learning, the correct answer may not be labeled for each training data. That is, for example, the training data in supervised learning for data classification can be data in which each of the training data is labeled with a category. The labeled training data is input to a neural network, and an error can be calculated by comparing the output (category) of the neural network with the label of the training data. As another example, in the case of unsupervised learning for data classification, an error can be calculated by comparing the input training data with the output of the neural network. The calculated error is backpropagated in the reverse direction (i.e., from the output layer to the input layer) in the neural network, and the connection weight value of each node in each layer of the neural network can be updated through backpropagation. The change amount of the connection weight value of each node to be updated can be determined by a learning rate.The neural network calculations on the input data and the backpropagation of errors constitute a learning cycle (epoch). The learning rate can be applied in different ways depending on the number of iterations of the neural network learning cycle. For example, in the early stages of neural network learning, the learning rate can be increased to improve efficiency by allowing the neural network to quickly achieve a certain level of performance, and in the later stages of learning, the learning rate can be decreased to improve accuracy.

[0061] In the training of a neural network, the training data may generally be a subset of the actual data (i.e., data to be processed using the trained neural network), and therefore there may be a training cycle in which the error associated with the training data decreases, but the error associated with the actual data increases. Overfitting is a phenomenon in which errors increase in the actual data due to excessive learning of the training data. For example, a neural network that has learned about cats by looking at a yellow cat may be unable to recognize that it is a cat when it sees a cat of a color other than yellow, which may be a type of overfitting. Overfitting may cause an increase in errors in machine learning algorithms. To prevent overfitting, various optimization methods may be applied. To prevent overfitting, methods such as increasing the training data, regularization, dropout, which deactivates some of the nodes of the network during the training process, and the use of a batch normalization layer may be applied.

[0062] FIG. 3 is a flow chart illustrating a process for training a graph neural network model according to one embodiment of the present disclosure. According to FIG. 3, a process of training a graph neural network model according to one embodiment of the present disclosure may include a step of determining nodes based on input variables of training data (S310), a step of determining edges connecting each node based on relationships between the input variables (S320), a step of generating a first adjacency matrix based on a graph composed of the nodes and edges (S330), and a step of training the graph neural network model based on the training data and the first adjacency matrix (S340).

[0063] In step S310, the processor 110 can determine a node based on the input variables of the learning data. In the present disclosure, one input variable included in the learning data can be determined as one node. For example, if the learning data is time-series data recording the medical data of a patient and includes five input variables {heart rate, respiration rate, systolic blood pressure, body temperature, SpO2}, the heart rate can be determined as the first node, the respiration rate as the second node, the systolic blood pressure as the third node, the body temperature as the fourth node, and the SpO2 value as the fifth node.

[0064] In step S320, the processor 110 may determine edges connecting each node based on the relationship between the input variables. For example, the learning data may include the first node to the fourth node, among which the first node and the second node, the first node and the third node, the first node and the fourth node, and the second node and the third node may be identified as input variables related to each other. In this case, the first node and the second node may be connected by an edge, the first node and the third node may be connected by an edge, the first node and the fourth node may be connected by an edge, and the second node and the third node may be connected by an edge.

[0065] In step S330, the processor 110 can generate a first adjacency matrix based on a graph composed of nodes and edges. In the present disclosure, each row and each column constituting the adjacency matrix can represent a respective node. Also, each element A of the adjacency matrix A ijcan be determined to be different values ​​depending on whether the i-th node and the j-th node are connected by an edge. For example, if the i-th node and the j-th node are connected by an edge, A ij = 1, if not connected by edge, A ij = 0. ii If the rows and columns have the same elements, such as ii =1.

[0066] When the graph in which the input variables of the learning data are represented by nodes and edges is as shown in FIG. 4, the first adjacency matrix can be expressed as shown in [Equation 1].

[0067]

number

[0068] In step S340, the processor 110 can train a graph neural network model based on the training data and the first adjacency matrix.

[0069] In one embodiment of the present disclosure, the graph neural network model may include a graph convolution network. In the graph convolution network, the weights included in each layer constituting the network may be updated as shown in [Equation 2].

[0070]

number

[0071] In this case, A is the first adjacent matrix, II (l) is the l-th feature matrix, W (l)may mean a matrix including l updated weights of a graph convolutional network. In this case, each weight constituting the matrix W may be a learnable parameter. The number of rows of the matrix W may be the same as the number of features of the previous layer in the graph convolutional network, and the number of columns may be the same as the number of features of the next layer.

[0072] The processor 110 can perform propagation to the network based on Equation 2.

[0073] FIG. 4 is a conceptual diagram illustrating training data of a graph neural network model according to an embodiment of the present disclosure.

[0074] 4 is composed of a total of four input variables, and each input variable is represented by a first node 410, a second node 420, a third node 430, and a fourth node 440. The first node 410 and the second node 420, the first node 410 and the third node 430, the first node 410 and the fourth node 440, and the second node 420 and the third node 430 are connected by edges.

[0075] The processor 110 may generate a first adjacency matrix based on whether each node is connected to other nodes by an edge. In the case of FIG. 4, the first adjacency matrix may be expressed as follows:

[0076]

number

[0077] The processor 110 may train a graph neural network model based on the data 411, 421, 431, and 441 for each input variable and the first adjacency matrix. As a result of training the graph neural network model, a matrix W representing the weight of the graph convolution network may be updated.

[0078] FIG. 5 is a conceptual diagram illustrating input data for a graph neural network model according to one embodiment of the present disclosure.

[0079] 5 is composed of three input variables, and each input variable is represented by a first node 510, a second node 520, and a third node 530. Furthermore, the first node 510 and the second node 520, the first node 510 and the third node 530, and the second node 520 and the third node 530 are connected by edges.

[0080] In the present disclosure, the graph neural network model to which the input data is input may be an artificial neural network model trained with training data including a total of four input variables including the input variables of the input data.

[0081] Conventionally, when generating prediction values ​​using a graph neural network model including a graph convolutional network, the adjacency matrix generated when the input data has three input variables is expressed as a matrix representing the relationship between the three input variables, as shown in [Mathematical Formula 4].

[0082]

number

[0083] In this case, in order to match the input data and morphology, only the relationships between the three input variables were learned during the model training process, and the weights of the graph convolutional network were updated; the relationships between the other input variables that were not included in the input data were not reflected in the training process.

[0084] In the present disclosure, the processor 110 may generate a second adjacency matrix based on whether each node is connected to other nodes by an edge. In this case, the processor 110 may generate a first adjacency matrix based on the relationship between input variables included in the training data of the graph neural network model, and then generate the second adjacency matrix by masking some elements.

[0085] For example, when the input variables included in the input data are represented by a first node 510, a second node 520, and a third node 530 as shown in Fig. 5, the processor 110 may generate a first adjacency matrix taking into consideration up to a fourth node 540, which is a node representing an input variable that is not included in the input data but is included in the learning data, and then may generate a second adjacency matrix by setting the value of an element associated with the fourth node 540 to 0. In this case, the second adjacency matrix may be expressed as follows:

[0086]

number

[0087] The processor 110 can generate a prediction value for the input data based on the respective data 511, 521, 531 corresponding to the input variables, the second adjacency matrix, and a matrix including the learned parameters, i.e., the updated weights.

[0088] In one embodiment of the present disclosure, the graph neural network model may include a graph attention network (GAT). In the present disclosure, the GAT is an artificial neural network that receives input data that can be represented in a graph format and generates an output, and may refer to an artificial neural network that can reflect the relative importance of each node constituting a graph by utilizing an attention method. For example, in one embodiment of the present disclosure, when generating a first adjacency matrix, the processor 110 may generate the first adjacency matrix based on the relationship between a first number of input variables and the relative importance between the first number of input variables.

[0089] Specifically, the processor 110 calculates each element A of the first adjacent matrix A by ij refers to whether the i-th node and the j-th node are connected by an edge and the relative importance of the i-th node and the j-th node determined using the attention method, and can be determined as a real value between 0 and 1.

[0090] In the present disclosure, when the attention method is utilized to generate the first adjacency matrix, the relative importance between input variables can be reflected in each parameter of the artificial neural network during learning, and therefore more accurate results can be generated during model inference after learning is completed.

[0091] Unlike the case where the graph neural network model is learned based on the relationship between limited input variables, the graph neural network model according to an embodiment of the present disclosure is a model that learns the relationship between input variables not included in the input data in addition to the input variables included in the input data. For example, the graph neural network may be learned based on learning data including four input variables, namely, heart rate, respiration rate, systolic blood pressure, and body temperature, and the input data may include only three input variables, namely, heart rate, respiration rate, and systolic blood pressure, in the inference stage of the graph neural network. In this case, the graph neural network generates an output value based on not only the relationship information between the heart rate, respiration rate, and systolic blood pressure, but also the relationship information between the heart rate, respiration rate, and systolic blood pressure when the body temperature not included in the input data is further taken into consideration.

[0092] Since the graph neural network model is trained based on training data that has richer relationship information than the input data, and the weights of the graph neural network model, i.e., the trainable parameters, are determined, the graph neural network model can generate more accurate prediction values.

[0093] In accordance with one embodiment of the present disclosure, a computer-readable storage medium having a data structure stored thereon is disclosed.

[0094] A data structure can refer to the organization, management, and storage of data that allows efficient access and modification of the data. A data structure can refer to the organization of data to solve a specific problem (e.g., data retrieval, data storage, data modification in the shortest time). A data structure can also be defined as a physical or logical relationship between data elements designed to support a specific data processing function. A logical relationship between data elements can include a connected relationship between data elements as a user thinks it should be. A physical relationship between data elements can include an actual relationship between data elements that is physically stored in a computer-readable storage medium (e.g., a hard disk).

[0095] Specifically, a data structure may include a collection of data, relationships between data, and functions or commands that can be applied to the data. Effectively designed data structures allow a computing device to perform calculations while minimizing the use of computing device resources. Specifically, a computing device can increase the efficiency of operations such as operations, reads, inserts, deletes, compares, swaps, and searches through effectively designed data structures.

[0096] Data structures can be classified into linear data structures and non-linear data structures according to the type of the data structure. A linear data structure can be a structure in which only one piece of data is linked after another. Linear data structures can include lists, stacks, queues, and deques. A list can refer to a series of data sets that are internally ordered. A list can include linked lists. A linked list can be a data structure in which data is linked in a manner in which each piece of data is linked in a row with a pointer. In a linked list, the pointer can contain information about the connection to the next or previous piece of data. A linked list can be expressed as a singly linked list, a doubly linked list, or a circularly linked list depending on the type of the data structure. A stack can be a data list structure that allows limited access to data. A stack can be a linear data structure in which data can be manipulated (e.g., inserted or deleted) only at one end of the data structure. Data stored in a stack can be a LIFO (Last in First Out) data structure in which the last in comes first. A queue is a data structure that allows limited access to data, and unlike a stack, it can be a FIFO (First in First out) data structure where the slowest data is stored is the slowest data is available. A deck can be a data structure that allows data to be processed on both ends of the data structure.

[0097] A nonlinear data structure may be a structure in which multiple pieces of data are concatenated after one piece of data. A nonlinear data structure may include a graph data structure. A graph data structure may be defined by vertices and edges, and the edges may include a line connecting two different vertices. A graph data structure may include a tree data structure. A tree data structure may be a data structure in which a path connecting two different vertices among multiple vertices included in a tree is one data structure. In other words, a graph data structure may be a data structure that does not form a loop.

[0098] Throughout this specification, the terms computational model, neural network, network function, and neural network can be used interchangeably. (Hereinafter, they will be uniformly described as neural network.) The data structure may include a neural network. And, the data structure including the neural network may be stored in a computer-readable storage medium. The data structure including the neural network may also include data input to the neural network, weights of the neural network, hyperparameters of the neural network, data acquired from the neural network, activation functions associated with each node or layer of the neural network, and loss functions for training the neural network. The data structure including the neural network may include any of the components disclosed above.

[0099] That is, the data structure including the neural network may include all or any combination of data input to the neural network, weights of the neural network, hyperparameters of the neural network, data acquired from the neural network, activation functions associated with each node or layer of the neural network, and loss functions for training the neural network. In addition to the above configurations, the data structure including the neural network may include any other information that determines the characteristics of the neural network. Also, the data structure may include all forms of data used or generated in the operation process of the neural network, and is not limited to the above. The computer-readable storage medium may include a computer-readable recording medium and / or a computer-readable transmission medium. A neural network may be composed of a collection of interconnected computational units, generally called nodes. Such nodes may be called neurons. A neural network may include at least one or more nodes.

[0100] The data structure may include data to be input to the neural network. The data structure including the data to be input to the neural network may be stored in a computer-readable storage medium. The data to be input to the neural network may include training data input during the training process of the neural network and / or input data to be input to the neural network after training has been completed. The data to be input to the neural network may include data that has undergone pre-processing and / or data to be pre-processed. The pre-processing may include a data processing process for inputting data to the neural network. Thus, the data structure may include data to be pre-processed and data generated in the pre-processing. The above data structures are merely examples, and the present disclosure is not limited thereto.

[0101] The data structure may include weights of the neural network. (In this specification, weights and parameters may be used interchangeably.) The data structure including the weights of the neural network may be stored in a computer-readable storage medium. The neural network may include a plurality of weights. The weights are variable and may be varied by a user or an algorithm to allow the neural network to perform a desired function. For example, when one or more input nodes are interconnected to an output node by respective links, the output node may determine an output node value based on values ​​input to the input nodes connected to the output node and parameters set for the links corresponding to the respective input nodes. The above data structures are merely examples, and the present disclosure is not limited thereto.

[0102] By way of example and not limitation, the weights may include weights that are variable during the neural network training process and / or weights at which neural network training has been completed. The weights that are variable during the neural network training process may include weights at the start of a learning cycle and / or weights that are variable during a learning cycle. The weights at which neural network training has been completed may include weights at which a learning cycle has been completed. Thus, a data structure including weights of a neural network may include a data structure including weights that are variable during the neural network training process and / or weights at which neural network training has been completed. Thus, the above-mentioned weights and / or combinations of each weight are included in a data structure including weights of a neural network. The above-mentioned data structures are merely exemplary and the present disclosure is not limited thereto.

[0103] The data structure including the neural network weights may be stored in a computer-readable storage medium (e.g., memory, hard disk) after a serialization process. Serialization may be a process of converting a data structure into a form that can be stored in the same or another computing device and later reconstructed and used. The computing device may serialize the data structure and transmit and receive the data over a network. The serialized data structure including the neural network weights may be reconstructed in the same or another computing device through deserialization. The data structure including the neural network weights is not limited to serialization. Furthermore, the data structure including the neural network weights may include a data structure for increasing the efficiency of calculations while minimizing the use of computing device resources (e.g., nonlinear data structures such as B-Tree, Trie, m-way search tree, AVLtree, Red-Black Tree). The above is merely an example, and the present disclosure is not limited thereto.

[0104] The data structure may include hyper-parameters of the neural network. The data structure including the hyper-parameters of the neural network may be stored in a computer-readable storage medium. The hyper-parameters may be variables that are changed by a user. The hyper-parameters may include, for example, a learning rate, a cost function, a number of learning cycle iterations, weight initialization (e.g., setting a range of weights to be initialized), and the number of Hidden Units (e.g., the number of hidden layers, the number of nodes in the hidden layers). The above data structures are merely examples, and the present disclosure is not limited thereto.

[0105] FIG. 10 is a simplified general schematic diagram of an exemplary computing environment in which embodiments of the present disclosure can be implemented.

[0106] Although the present disclosure has been described above as generally being embodied in a computing device, those skilled in the art will appreciate that the present disclosure may also be embodied in combination with computer-executable instructions and / or other program modules capable of being executed on one or more computers and / or as a combination of hardware and software.

[0107] Generally, modules herein include routines, programs, components, data structures, etc. that perform particular tasks or implement particular abstract data types. Additionally, those skilled in the art will appreciate that the methods of the present disclosure can be practiced with other computer system configurations, including single-processor or multi-processor computer systems, minicomputers, mainframe computers, as well as personal computers, handheld computing devices, microprocessor-based or programmable consumer electronics, etc., each of which can operate in conjunction with one or more associated devices.

[0108] The embodiments described in this disclosure may also be practiced in distributed computing environments where certain tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote memory storage devices.

[0109] A computer includes a variety of computer-readable media. Any medium accessible by a computer can be a computer-readable medium, including volatile and non-volatile media, transitory and non-transitory media, mobile and non-mobile media. By way of example and not limitation, computer-readable media can include computer-readable storage media and computer-readable transmission media. Computer-readable storage media include volatile and non-volatile media, transitory and non-transitory media, mobile and non-mobile media implemented in any method or technology for storing information such as computer-readable instructions, data structures, program modules or other data. Computer-readable storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, DVD (digital video disk) or other optical disk storage devices, magnetic cassettes, magnetic tapes, magnetic disk storage devices or other magnetic storage devices, or any other medium that can be accessed by a computer and used to store information.

[0110] Computer-readable transmission media typically embodies computer-readable instructions, data structures, program modules or other data in a modulated data signal, such as a carrier wave or other transport mechanism, and includes all information delivery media. The term modulated data signal means a signal that has one or more of its characteristics set or changed in such a manner as to encode information in the signal. By way of example, and not limitation, computer-readable transmission media include wired media, such as a wired network or direct-wired connection, and wireless media, such as acoustic, RF, infrared and other wireless media. Any combination of any of the above media should also be included within the scope of computer-readable transmission media.

[0111] An exemplary environment (1100) for implementing various aspects of the present disclosure is shown including a computer (1102) including a processing unit (1104), a system memory (1106), and a system bus (1108). The system bus (1108) couples system components, including but not limited to the system memory (1106), to the processing unit (1104). The processing unit (1104) can be any of a variety of commercially available processors. Dual processors and other multi-processor architectures can also be utilized as the processing unit (1104).

[0112] The system bus (1108) can be any of several types of bus structures that can be further interconnected to a memory bus, a peripheral bus, and a local bus using any of a variety of commercially available bus architectures. The system memory (1106) includes a read only memory (ROM) (1110) and a random access memory (RAM) (1112). A basic input / output system (BIOS) is stored in non-volatile memory (1110), such as a ROM, EPROM, or EEPROM, and contains the basic routines that support the exchange of information between the various components within the computer (1102), such as during start-up. The RAM (1112) can also include high-speed RAM, such as static RAM, for caching data.

[0113] The computer 1102 also includes an internal hard disk drive (HDD) 1114 (e.g., EIDE, SATA) - although the internal hard disk drive 1114 can also be configured for external use in a suitable chassis (not shown) - a magnetic floppy disk drive (FDD) 1116 (e.g., for reading from and writing to a removable diskette 1118), and an optical disk drive 1120 (e.g., for reading from CD-ROM disks 1122 and for reading from and writing to other high capacity optical media such as DVDs). The hard disk drive 1114, magnetic disk drive 1116, and optical disk drive 1120 can be connected to the system bus 1108 by a hard disk drive interface 1124, a magnetic disk drive interface 1126, and an optical drive interface 1128, respectively. The interface (1124) for implementing an external drive includes, for example, at least one or both of the Universal Serial Bus (USB) and IEEE 1394 interface technologies.

[0114] These drives and their associated computer-readable media provide non-volatile storage of data, data structures, computer-executable instructions, and the like. In the case of computer 1102, the drives and media accommodate the storage of any data in a suitable digital format. While the above description of computer-readable storage media refers to HDDs, removable magnetic disks, and removable optical media such as CDs or DVDs, those skilled in the art will appreciate that other types of computer-readable storage media, such as zip drives, magnetic cassettes, flash memory cards, cartridges, and the like, may also be used in the exemplary operating environment, and further that any such media may contain computer-executable instructions for performing the methods of the present disclosure.

[0115] A number of program modules may be stored in the drives and RAM 1112, including an operating system 1130, one or more application programs 1132, other program modules 1134, and program data 1136. All or portions of the operating system, applications, modules, and / or data may also be cached in RAM 1112. It will be appreciated that the present disclosure may be implemented with various commercially available operating systems or combinations of operating systems.

[0116] A user may enter commands and information into the computer 1102 through one or more wired or wireless input devices, such as a keyboard 1138 and a pointing device, such as a mouse 1140. Other input devices (not shown) may include a microphone, an IR remote control, a joystick, a game pad, a stylus pen, a touch screen, and the like. These and other input devices are often connected to the processing unit 1104 through an input device interface 1142 that is connected to the system bus 1108, but may also be connected through other interfaces, such as a parallel port, an IEEE 1394 serial port, a game port, a USB port, an IR interface, and the like.

[0117] A monitor 1144 or other type of display device is also connected to the system bus 1108 through an interface, such as a video adapter 1146. In addition to the monitor 1144, computers typically include other peripheral output devices (not shown), such as speakers, printers, etc.

[0118] The computer 1102 may operate in a networked environment using logical connections to one or more remote computers, such as remote computer(s) 1148, via wired and / or wireless communications. The remote computer(s) 1148 may be a workstation, a server computer, a router, a personal computer, a handheld computer, a microprocessor-based entertainment device, a peer device or other conventional network node, and typically includes many or all of the components depicted for the computer 1102, although for simplicity, only a memory storage device 1150 is illustrated. The logical connections depicted include wired and wireless connections in a local area network (LAN) 1152 and / or larger networks, such as a wide area network (WAN) 1154. Such LAN and WAN networking environments are commonplace in offices and companies and facilitate enterprise-wide computer networks, such as intranets, all of which may connect to a global computer network, such as the Internet.

[0119] When used in a LAN networking environment, the computer 1102 connects to the local network 1152 through a wired and / or wireless communication network interface or adapter 1156. The adapter 1156 can facilitate wired or wireless communication to the LAN 1152, which may also include a wireless access point installed thereon for communicating with the wireless adapter 1156. When used in a WAN networking environment, the computer 1102 can include a modem 1158 or have other means for establishing communications over the WAN 1154, such as connecting to a communications server on the WAN 1154 or through the Internet. The modem 1158, which may be internal or external and may be a wired or wireless device, connects to the system bus 1108 through the serial port interface 1142. In a networked environment, program modules described for computer 1102, or portions thereof, may be stored in a remote memory / storage device 1150. It will be readily appreciated that the network connections shown are exemplary and other means of establishing a communications link between the computers may be used.

[0120] The computer (1102) is operable to communicate with any wireless device or unit that is deployed and operates in wireless communication, such as printers, scanners, desktop and / or handheld computers, portable data assistants (PDAs), communication satellites, any equipment or location associated with a wirelessly detectable tag, and telephones. This includes at least Wi-Fi and Bluetooth wireless technologies. Thus, communication can be in a predefined structure, such as a traditional network, or simply ad hoc communication between at least two devices.

[0121] Wi-Fi (Wireless Fidelity) allows devices to connect to the Internet without being wired. Wi-Fi is a wireless technology like a cell phone that allows such devices, e.g., computers, to send and receive data indoors and outdoors, anywhere within the range of a base station. Wi-Fi networks use IEEE 802.11 (a, b, g, etc.) radio technology to provide secure, reliable, and fast wireless connections. Wi-Fi can be used to connect computers to each other, to the Internet, and to wired networks (using IEEE 802.3 or Ethernet). Wi-Fi networks can operate in the unlicensed 2.4 and 5 GHz radio bands, e.g., at data rates of 11 Mbps (802.11a) or 54 Mbps (802.11b), or in products that include both bands (dual bands).

[0122] Those of ordinary skill in the art will appreciate that information and signals may be represented using any of a variety of different technologies and techniques. For example, the data, instructions, commands, information, signals, bits, symbols, and chips referred to in the above description may be represented by voltages, currents, electromagnetic waves, magnetic fields, etc. or particles, optical fields, etc. or particles, or any combination thereof.

[0123] Those of ordinary skill in the art will appreciate that the various exemplary logic blocks, modules, processors, means, circuits, and algorithm steps described in the description of the embodiments disclosed herein can be implemented by electronic hardware, various forms of program or design code (for convenience, referred to herein as "software"), or a combination of all of these. To clearly illustrate such interoperability of hardware and software, the various exemplary components, blocks, modules, circuits, and steps have been generally described above with a focus on their functionality. Whether such functionality is implemented in hardware or software depends on the design constraints imposed on the particular application and the overall system. Those of ordinary skill in the art will appreciate that the described functionality can be implemented in various ways for each particular application, and such implementation decisions should not be interpreted as departing from the scope of the present disclosure.

[0124] The various embodiments described herein may be implemented by a method, apparatus, or article of manufacture using standard programming and / or engineering techniques. The term "article of manufacture" includes a computer program, carrier, or media accessible by any computer-readable device. For example, computer-readable storage media include, but are not limited to, magnetic storage devices (e.g., hard disks, floppy disks, magnetic strips, etc.), optical disks (e.g., CDs, DVDs, etc.), smart cards, and flash memory devices (e.g., EEPROM, cards, sticks, key drives, etc.). Additionally, the various storage media described herein may include one or more devices and / or other machine-readable media for storing information.

[0125] It should be understood that the particular order or hierarchy of the stages in the processes depicted is an example of an exemplary approach. Based on design priorities, it should be understood that the particular order or hierarchy of the stages in the processes can be rearranged within the scope of the present disclosure. The accompanying method claims provide elements of the various stages in a sample order, but are not meant to be limited to the particular order or hierarchy depicted.

[0126] The description of the illustrated embodiments is provided to enable any person of ordinary skill in the art to which the disclosure pertains to the present disclosure to utilize or practice the present disclosure. Various modifications to such embodiments will be apparent to those of ordinary skill in the art to which the disclosure pertains, and the general principles defined herein may be applied to other embodiments without departing from the scope of the present disclosure. Thus, the present disclosure is not intended to be limited by the embodiments shown herein, but is to be accorded the broadest scope consistent with the principles and novel features disclosed herein. [Explanation of symbols]

[0127] 100: Computing equipment 110: Processor 130: Memory 150: Network section 410: First node 411: Data 420: Second node 421: Data 430: Third node 431: Data 440: 4th node 441: Data 510: First node 511: Data 520: Second node 521: Data 530: Third node 531: Data 540: 4th node

Claims

1. 1. A method, performed via a computing device, for making predictions utilizing an artificial neural network, comprising: obtaining input data including a second number of input variables; and generating a predicted value based on a second adjacency matrix generated based on relationships between the second number of input variables and inputting the input data into a graph neural network model; Including, The graph neural network model is the artificial neural network model trained based on a first adjacency matrix generated based on relationships between a first number of input variables, the first number being greater than or equal to the second number.

2. 10. The method of claim 1 , The graph neural network model is an artificial neural network model trained using training data including the first number of input variables; method.

3. 3. The method of claim 2, The graph neural network model is generating the first adjacency matrix based on relationships between the first number of input variables; and training the graph neural network model based on the training data and the first adjacency matrix; This is an artificial neural network model trained with method.

4. 10. The method of claim 1 , generating a predicted value based on inputting data for the second number of input variables into a graph neural network model, generating the second adjacency matrix based on relationships between the second number of input variables; and generating a prediction value for the input data based on the input data, the second adjacency matrix, and learned parameters of the graph neural network model; Including, method.

5. 5. The method of claim 4, generating the second adjacency matrix based on a relationship between input variables of the input data, generating a first adjacency matrix based on relationships between the first number of input variables; and masking elements of the first adjacency matrix that do not represent relationships between the second number of input variables; Including, method.

6. 6. The method of claim 5, The step of generating a first adjacency matrix based on relationships between the first number of input variables includes: generating the first adjacency matrix based on relationships between and relative importance between the first number of input variables; Including, method.

7. 4. The method of claim 3, Each of the first adjacency matrix and the second adjacency matrix is A matrix generated based on a graph composed of nodes and edges. method.

8. 8. The method of claim 7, The graph is determining nodes based on input variables; and determining edges connecting each of the nodes based on relationships between the input variables; This is a graph generated by including method.

9. 10. The method of claim 1 , The graph neural network model is An artificial neural network model that takes a patient's time-series medical data as input and outputs a diagnosis; Including, method.

10. 1. A computer program stored on a computer readable storage medium comprising operations for generating, via a computing device, a data set associated with a prediction of a state of a subject, the operations comprising: obtaining input data including a second number of input variables; and generating a second adjacency matrix based on relationships between the second number of input variables, and generating a predicted value based on inputting the input data into a graph neural network model; Including, The graph neural network model is a computer program stored on a computer-readable storage medium, the computer program being an artificial neural network model trained based on a first adjacency matrix generated based on relationships between a first number of input variables, the first number being greater than or equal to the second number.

11. 1. A computing device comprising: one or more processors; and Memory; Including, The processor, Obtaining input data including a second number of input variables; and generating a predicted value based on a second adjacency matrix generated based on relationships between the second number of input variables and inputting the input data into a graph neural network model; The graph neural network model is a first adjacency matrix generated based on relationships between a first number of input variables, the first number being greater than or equal to the second number;

Images