Model learning device, model learning method, and model learning program

Abstract

This model learning device comprises: an input data processing unit that acquires input data; a set parameter processing unit that acquires a set parameter; a model parameter learning unit that receives input of the input data and the set parameter to learn to rank model parameters by using an objective function that includes a loss function holding an order relation and a regularization term for limiting output of a model; and a model parameter processing unit that outputs the learned model parameters. The input data is observable multivariate data and label data indicating an order relation. The set parameter includes a hyperparameter of the regularization term. The model parameter learning unit obtains model parameters that result in output of a score close to an objective variable.

Description

[Technical field]

[0001] The present invention relates to a model learning device, a model learning method, and a model learning program. [Background technology]

[0002] Learning a model that represents input-output relationships from multivariate data is one of the problems in the fields of machine learning and artificial intelligence. In this problem, in a typical setting, a set of pairs of input values ​​and output values ​​obtained from those inputs is given as numerical data. However, when responses cannot be obtained from questionnaires due to psychological resistance, or when output values ​​are difficult to observe, ordinal relationships rather than numerical values ​​may be given as output values.

[0003] Examples of cases where values ​​cannot be obtained due to psychological resistance include the collection and analysis of physical information or information about private life. For example, when conducting a questionnaire survey to investigate the relationship between diet and health, people may not provide their weight due to psychological resistance. In other cases, data may be collected to predict electricity consumption from indoor conditions (temperature, humidity, etc.). Here too, actual values ​​may not be provided due to psychological resistance, such as not wanting people to guess how long they spend at home or their family composition.

[0004] In addition, examples of output values ​​that are difficult to observe include emotions such as satisfaction and excitement. In general, the true values ​​of human emotions cannot be observed, and absolute evaluation is difficult because they include the subjectivity of the evaluator. For this reason, data on human emotions is often obtained by multi-stage evaluation such as Likert scales (Non-Patent Document 1) and SAM (self-assessment manikins) (Non-Patent Document 2). These evaluation scales are non-linear and include the bias of the evaluator, so it is not appropriate to consider them as interval scales or proportional scales. In fact, it is not appropriate to compare an emotion evaluation value of 2 points by a certain evaluator A with an emotion evaluation value of 4 points by evaluator B and argue that there is a difference of two times, or to interpret the same difference of 2 points as being equivalent to an evaluation result of 1 point and 3 points.

[0005] These problems are alleviated by making the questions in the questionnaire ordinal scales or interpreting the answers as ordinal scales. Making the questions ordinal scales means asking questions such as "Is your weight more than ~ kg?" instead of "Please enter your weight." By not referring to specific numerical values, it is possible to reduce psychological resistance and make it easier to obtain answers. Also, in the case of emotions and feelings, from the emotional evaluation value of 2 points for evaluator A and the emotional evaluation value of 4 points for evaluator B, it can be interpreted as "A < B." This "A < B" treats the answer as an order relation, and compared with the case of treating it as a numerical value, the possibility of including incorrect interpretations can be reduced. From such a setting, although specific numerical values cannot be obtained, when an order relation is given as label data, it is necessary to consider the problem of predicting specific numerical values from that order relation.

[0006] There is rank learning as a method of learning model parameters from order relations. Based on the commonly used pairwise rank learning, the general flow of rank learning is described below. For n k , k observable multivariate data X with elements, a set D of n k pairs indicating order relations is defined as label data. The multivariate data X and the set D of pairs are as follows.

[0007]

Number

[0008] This does not indicate v k > u k but rather that when the input values are v k , u k the target variables y vk , y uk are such that y vk > y uk . And if the model to be learned is f, then for this v k , u k the output values f(v k ), f(u k ) are f(vk )>f(u k ) so that L=exp(f(v k )-f(u k Learning is achieved by minimizing . In other words, the goal of learning to rank is to adjust the parameters of the score output model f to construct the intended ranking.

[0009] A well-known rank learning method is RankNet (Non-Patent Document 3), which uses a logistic function and an entropy loss function, and defines the loss function L as follows:

[0010]

number

[0011] This loss function produces smaller output when the order relationship of paired data is maintained, making it possible to generate a score that satisfies the order relationship of label data.

[0012] There is also a rank learning method using Gaussian distribution (Non-Patent Document 4). In this method, the loss function L is defined as follows:

[0013]

number

[0014] These methods adjust model parameters based on the order relationship and generate a score that can be used to construct a ranking. In other words, the value of the score does not matter as long as a ranking based on the label data can be constructed. For example, the input v k ,u k In contrast, u k than v k By learning from this label, f(v k )>f(u k ) is possible, and a ranking based on the label data can be constructed. However, f(v k),f(u k ) is not guaranteed to be in the closed interval [1,5], and this score cannot be considered as an emotional assessment value. [Prior art documents] [Non-patent literature]

[0015] [Non-Patent Document 1] R. Likert. A technique for the measurement of attitudes. Archives of Psychology, Vol. 22 140, pp. 5-55, 1932. [Non-Patent Document 2] Jon D. Morris. Observations: SAM: The Self-Assessment Manikin An Efficient Cross-Cultural Measurement Of Emotional Response. (This article originally appeared in the Journal of Advertising Research November / December 1995.) [Non-Patent Document 3] Christopher Burges, Tal Shaked, Erin Renshaw, Ari Lazier, Matt Deeds, Nicole Hamilton, and Greg Hullender. Learning to rank using gradient descent. pp. 89-96, 01 2005. [Non-Patent Document 4] Wei Chu and Zoubin Ghahramani. Preference learning with gaussian processes. pp. 137-144, 08 2005. Summary of the Invention [Problem to be solved by the invention]

[0016] However, existing rank learning is a model that learns order relationships from label data and outputs scores according to the order relationships, but the output does not indicate the objective variable. In other words, when only the order relationships are given as label data, it is difficult to predict a specific objective variable from the order relationships using existing rank learning alone.

[0017] The present invention has been made in light of the above-mentioned circumstances, and its purpose is to provide a model learning device, a model learning method, and a model learning program that predict a dependent variable from an order relationship even when only an order relationship is obtained as label data. [Means for solving the problem]

[0018] One aspect of the present invention is a model learning device. The model learning device includes an input data processing unit that acquires input data, a setting parameter processing unit that acquires setting parameters, a model parameter learning unit that receives the input data and the setting parameters and rank-learns model parameters using an objective function composed of a loss function that maintains an order relationship and a regularization term that limits the output of a model, and a model parameter processing unit that outputs the learned model parameters. The input data is observable multivariate data and label data that indicates an order relationship. The setting parameters include a hyperparameter of the regularization term. The model parameter learning unit rank-learns model parameters using an objective variable Predict Get the model parameters that will output the score.

[0019] One aspect of the present invention is The computer runsThis is a model learning method. The model learning method includes acquiring input data, acquiring setting parameters, and using the input data and the setting parameters as inputs, rank learning of model parameters using an objective function composed of a loss function that maintains an order relationship and a regularization term that limits the output of a model, and outputting the learned model parameters. The input data is observable multivariate data and label data that indicates an order relationship. The setting parameters include a hyperparameter of the regularization term. The rank learning includes: Predict Get the model parameters that will output the score.

[0020] A model learning program according to one aspect of the present invention causes a computer to execute the functions of each of the components of the model learning device described above. Effect of the Invention

[0021] According to the present invention, even when only an order relationship is obtained as label data, a model learning device, a model learning method, and a model learning program are provided that predict a dependent variable from the order relationship. [Brief description of the drawings]

[0022] [Figure 1] FIG. 1 is a block diagram illustrating an example of a functional configuration of a model learning device according to an embodiment. [Diagram 2] FIG. 2 is a block diagram illustrating an example of a hardware configuration of the model learning device according to the embodiment. [Diagram 3] FIG. 3 is a flowchart illustrating an example of a processing procedure and processing contents of model parameter estimation executed by the model learning device according to the embodiment. [Figure 4] FIG. 4 is a flowchart illustrating an example of a processing procedure and processing contents of model parameter rank learning executed by the model parameter learning unit of the model learning device according to the embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0023] Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

[0024] [Configuration example] First, the operation of the model learning device 1 will be described with reference to Fig. 1. Fig. 1 is a block diagram showing an example of the functional configuration of the model learning device 1 according to the embodiment.

[0025] 1, the model learning device 1 has an input data processing unit 10, a setting parameter processing unit 20, a model parameter learning unit 30, a model parameter processing unit 40, a recording unit 50, and an input / output unit 60. The recording unit 50 has an input data recording unit 51, a setting parameter recording unit 52, and a model parameter recording unit 53.

[0026] The input / output unit 60 inputs and outputs data to and from the external device 2. The input / output unit 60 receives input data and setting parameters from the external device 2. In addition, the input / output unit 60 outputs model parameters to the external device 2.

[0027] The input data processing unit 10 acquires input data from the external device 2 via the input / output unit 60. The input data processing unit 10 outputs the acquired input data to the input data recording unit 51.

[0028] The input data recording unit 51 receives the input data from the input data processing unit 10 and records it.

[0029] The setting parameter processing unit 20 acquires setting parameters from the external device 2 via the input / output unit 60. The setting parameter processing unit 20 outputs the acquired setting parameters to the setting parameter recording unit 52.

[0030] The setting parameter recording unit 52 receives the setting parameters from the setting parameter processing unit 20 and records them.

[0031] The model parameter learning unit 30 receives input data recorded in the input data recording unit 51 and setting parameters recorded in the setting parameter recording unit 52, and performs rank learning of the model parameters using an objective function consisting of a loss function that maintains the order relationship and a regularization term that limits the output of the model. The model parameter learning unit 30 outputs the learned model parameters to the model parameter recording unit 53.

[0032] The model parameter recording unit 53 receives the model parameters from the model parameter learning unit 30 and records them.

[0033] The model parameter processing unit 40 reads the model parameters from the model parameter recording unit 53 , and outputs them to the external device 2 via the input / output unit 60 .

[0034] The input data acquired by the input data processing unit 10 is observable multivariate data and label data indicating an order relationship. For example, the label data is a set of pairs indicating an order relationship.

[0035] The setting parameters acquired by the setting parameter processing unit 20 include hyperparameters of the regularization term of the objective function used by the model parameter learning unit 30 for rank learning. The hyperparameters include quantiles of the objective variable. Here, a quantile is a number that exists at a position where a set of sorted numbers is equally divided by positive integers. The hyperparameters also include parameters that determine the weight of the regularization term. The setting parameters also include a learning rate parameter used to optimize the objective function.

[0036] Next, we will explain the hardware configuration of the model learning device 1. The model learning device 1 is configured with a computer. For example, the model learning device 1 is configured with a personal computer, a server computer, or the like.

[0037] 2 is a block diagram showing an example of a hardware configuration of the model learning device 1 according to the embodiment. As shown in FIG. 2, the model learning device 1 includes an input / output interface 110, a CPU 120, and a storage device .

[0038] The input / output interface 110 , the CPU 120 , and the storage device 130 are electrically connected to one another via a bus 140 , and exchange data and commands via the bus 140 .

[0039] The input / output interface 110 is connected to the external device 2 via a signal cable or a network. The input / output interface 110 is used to receive data from the external device 2 and to output model parameters to the external device 2.

[0040] The storage device 130 stores programs and data necessary for processing executed by the CPU 120. The CPU 120 reads out the necessary programs and data from the storage device 130 and executes them to perform various types of processing.

[0041] The storage device 130 includes a main storage device 131 and an auxiliary storage device 132. The main storage device 131 and the auxiliary storage device 132 exchange programs and data with each other.

[0042] The main storage device 131 stores programs and data temporarily required for processing by the CPU 120. For example, the main storage device 131 is configured with a volatile memory such as a RAM (Random Access Memory).

[0043] The auxiliary storage device 132 stores programs and data supplied via an external device or a network, and provides the programs and data temporarily required for processing by the CPU 120 to the main storage device 131. For example, the auxiliary storage device 132 is configured with a non-volatile memory such as an HDD (Hard Disk Drive) or an SSD (Solid State Drive).

[0044] The CPU 120 is a processor, and is hardware that processes data and instructions. The CPU 120 includes a control device 121 and an arithmetic unit 122.

[0045] The control device 121 controls the input / output interface 110 , the arithmetic device 122 , and the storage device 130 .

[0046] The arithmetic unit 122 reads a program and data from the main memory device 131 , executes the program to process the data, and provides the processed data to the main memory device 131 .

[0047] In such a hardware configuration, the input / output interface 110 constitutes the input / output unit 60. The CPU 120 and the main memory device 131 constitute the input data processing unit 10, the setting parameter processing unit 20, the model parameter learning unit 30, and the model parameter processing unit 40. The memory device 130 (for example, the auxiliary memory device 132) constitutes the recording unit 50.

[0048] [Example] Next, the operation of the model learning device 1 will be described. The rank learning performed by the model learning device 1 is a method of restricting the distribution of output values ​​in rank learning by a regularization term based on quantiles. In order to bring the model output closer to the objective variable, it is necessary to approximate the model output distribution to the objective variable distribution. However, when only an order relationship is obtained as label data, a specific numerical value cannot be obtained, and therefore the objective variable distribution cannot be obtained.

[0049] Therefore, instead of approximating the entire distribution, we give the quantiles of the objective variable as hyperparameters and approximate only the multiple points determined by the quantiles. A quantile is a set of m-1 numbers {xi1,xi2,…,xi m-1}, and xi l is called the lmth quantile.

[0050] When the m-quantile α of the objective variable is given as a hyperparameter, the model output is restricted by minimizing the error between the m-quantile β of the model output value. Alternatively, the model output is restricted by minimizing an objective function that includes a term calculated using the m-quantile α of the objective variable and the m-quantile β of the model output value.

[0051] (Input data) The input data acquired by the input data processing unit 10 is observable multivariate data and label data indicating an order relationship. For example, the multivariate data has n elements. x The number of observable multivariate data X is n, and the label data indicates the order relationship. k Hereinafter, the set of pairs D is also referred to as paired label data D. The multivariate data X and paired label data D are expressed as follows.

[0052]

number

[0053] (Model) The model f used in the rank learning executed by the model parameter learning unit 30 can be any model that outputs a score that serves as a ranking construction index in rank learning. The input model f may be a linear model or a generalized linear model using a logistic function or a sigmoid function, or may be a nonlinear model such as a Gaussian process model or a deep learning model. In addition, when X is multivariate time series data, the latent variable z t-1 and x t From z t and the function to obtain z t to the objective variable y t It may be a basic time series model that estimates x, or an RNN architecture such as LSTM. In the case where the multivariate data X is a set of images, it is also assumed to use a model with a CNN structure expressed by the following equation.

[0054]

number

[0055] (output) The output of the model learning device 1 is an estimation result of the model parameter θ. The estimation result of the model parameter θ is a score close to the objective variable.

[0056] (Objective function) Next, we explain the objective function in rank learning. The objective function consists of a loss function E that maintains the order relationship and a regularization term Ω that limits the model output. Based on this, the objective function L is expressed as follows:

[0057]

number

[0058] Here, λ is a hyperparameter that determines the weight of the regularization term, and α is a hyperparameter that indicates the m-quantile of the objective variable.

[0059] (regularization) Since the subject of the objective function is regularization, we first discuss the regularization term Ω. There are various possible definitions for Ω, which approximates the quantiles of two distributions. Here, we define Ω1 and Ω2 as follows.

[0060] First, let us consider Ω1. If the cumulative distribution function of the objective variable is Φ(x) and the cumulative distribution function of the model output is Ψ(x), the L1 distance between the two cumulative distribution functions can be defined as a regularization term that takes a small value as the two distributions become more similar.

[0061]

number

[0062] In this formula, Φ(a) cannot be calculated because the specific value of the objective variable cannot be calculated. Therefore, the m-quantile of the objective variable, α={α1,…,α m-1} is given, and the following equation, in which the left-hand side of the previous equation is discretized with respect to a, is minimized.

[0063]

number

[0064] Here, the model output sorted in ascending order is y, and α l Let y be the subset of elements smaller than αl Then, Ψ(α l ) can be calculated as follows:

[0065]

number

[0066] However, n used in equation (7) is a counting process of the number of elements, which is procedural and generally cannot be used for backpropagation. Therefore, the (1 / m) elements belonging to the l+1th quantile of the model output y are calculated by α l From α l+1 By defining the regularization term Ω2 so that it falls within the range of Ψ(α l ) → l / m and approximately minimize Ω1. Ω2 is shown in equation (8).

[0067]

number

[0068] Here, b is the batch size (=n(y)), and α0,α m are arbitrary values ​​that indicate the upper and lower limits of the objective variable. This regularization term Ω2 divides the lm-th quantile of y by β l Then, the two quantiles (β l ,β l+1 ) element y i Whereas, y i is in the interval [α l ,α l+1 ], the value is 0, and y i is α l If smaller than αl The difference between i is α l+1 If greater than α l+1 This outputs the difference between the quantiles of y and the interval [β l ,β l+1 ] are in the interval [α l ,α l+1 ] means that the loss is smaller. Therefore, when the elements of y are equally divided into each interval determined by the quantile α, Ω2 = 0. In other words, when b is sufficiently large and Ω2 → 0, n(y αl )→(l / m)b, so by substituting into equations (5) and (7), we also get Ω→0. Therefore, minimizing Ω2 achieves the minimization of Ω1.

[0069] Next, we will explain the minimization algorithm of formula (8). First, for elements belonging to the l+1th quantile of the model output y, we define vector c that stores the (l+1)mth quantile of the objective variable and vector d that stores the lmth quantile of the objective variable as follows.

[0070]

number

[0071] Using these u and l, the regularization term for calculation processing is expressed as follows:

[0072]

number

[0073] In this regularization term, the hyperparameter α can be set to any value and can be changed according to the nature of the batch input during training. This is done by dividing a training data set into X1, X2, …, X nThis shows that when training by dividing the dataset into mini-batches with different properties, the hyper-parameters can be varied at the mini-batch level, such as hyper-parameter α for mini-batch X1 and hyper-parameter β for mini-batch X2. For example, in the task of predicting an individual's weight from their diet, it is possible to set hyper-parameters separately for when the input batch is male and when it is female. Also, in the task of predicting human emotions while watching a sporting event, it is possible to set hyper-parameters separately for exciting and non-exciting times.

[0074] The loss function E can be any function that reduces the output when maintaining the order of paired data. For example, it can be equation (1) as in RankNet (Non-Patent Document 3). It can also be equation (2) as in the rank learning method using Gaussian distribution (Non-Patent Document 4).

[0075] (Optimization method) Any optimization method can be applied to optimize the objective function, such as the gradient method, stochastic gradient method, Adam, etc. When using the gradient method, it is sufficient to repeatedly update the parameters according to the following formula at the kth optimization step.

[0076]

number

[0077] Here, γ k is the learning rate parameter. The gradient of the objective function ∇ θ L(θ) may be a function derived by calculation, or may be calculated numerically.

[0078] (Model parameter estimation) Next, the procedure and contents of the model parameter estimation process executed by the model learning device 1 will be described with reference to Fig. 3. Fig. 3 is a flow chart showing the procedure and contents of the model parameter estimation process executed by the model learning device 1.

[0079] In step S1, the input data processing unit 10 acquires input data. The input data is the above-mentioned multivariate data X and pair label data D. The input data processing unit 10 also stores the acquired input data in the input data recording unit 51.

[0080] In step S2, the setting parameter processing unit 20 acquires setting parameters. The setting parameters include hyperparameters of the regularization term of the objective function. The hyperparameters include the m-quantile α of the objective variable and a parameter λ that determines the weight of the regularization term. The setting parameters also include a learning rate parameter γ k The setting parameter processing unit 20 also stores the acquired setting parameters in the setting parameter recording unit 52.

[0081] In step S3, the model parameter learning unit 30 performs rank learning of the model parameters using an objective function having a loss function and a regularization term, with the input data recorded in the input data recording unit 51 and the setting parameters recorded in the setting parameter recording unit 52 as input. The model parameter learning unit 30 also stores the learned model parameters in the model parameter recording unit 53.

[0082] In step S 4 , the model parameter processing unit 40 reads the model parameters from the model parameter recording unit 53 , and outputs them to the external device 2 .

[0083] (Rank learning) Next, a process procedure and process contents of model parameter rank learning executed by the model parameter learning unit 30 will be described with reference to Fig. 4. Fig. 4 is a flowchart showing an example of a process procedure and process contents of model parameter rank learning executed by the model parameter learning unit 30.

[0084] In step S11, the model parameter θ is initialized.

[0085] In step S12, the maximum number of iterations of rank learning is set. Also, the number of calculation iterations of rank learning is initialized. That is, the number of calculation iterations is set to 0.

[0086] In step S13, the model parameter θ is updated according to equations (3) and (10).

[0087] In step S14, the number of calculation iterations is updated, that is, the number of calculation iterations is incremented by one.

[0088] In step S15, it is determined whether the number of calculation iterations exceeds the maximum number of iterations. If the number of calculation iterations does not exceed the maximum number of iterations, the process returns to step S13. If the number of calculation iterations exceeds the maximum number of iterations, learning of the model parameters is terminated, and the process returns to step S4 in FIG. 3.

[0089] [effect] According to the embodiment, even if only an order relationship is obtained as label data, it is possible to predict the objective variable from the order relationship. This makes it possible to learn an arbitrary rank learning model not only as a ranking generation model but also as a regression model that outputs a numerical value that approximates the distribution of the objective variable. This makes it possible to obtain not only an order relationship but also an estimated value as a numerical value for objective variables that are difficult to evaluate absolutely, such as emotions, when a survey response cannot be given in a specific numerical value due to psychological resistance.

[0090] In the above embodiment, an example is shown in which a gradient method is used during optimization, but any method such as a stochastic gradient method or Adam can be used. Similarly, any loss function can be used for E in the objective function (Equation 3). In the model learning device 1 shown in FIG. 1 of the above embodiment, the operation of each component can be constructed as a program, and the program can be installed and executed on a computer used as the model learning device, or distributed via a network. The present invention is not limited to the above embodiment, and various modifications and applications are possible.

[0091] For example, the above embodiment is an example of applying the present invention to a pairwise rank learning method, but the present invention is not limited to this and may be applied to other rank learning methods, such as a pointwise rank learning method or a listwise rank learning method.

[0092] The present invention is not limited to the above-mentioned embodiment, and can be modified in various ways without departing from the gist of the present invention. The embodiments may be combined as appropriate, and in that case, the combined effect can be obtained. Furthermore, the above-mentioned embodiment includes various inventions, and various inventions can be extracted by combinations selected from the multiple components disclosed. For example, if the problem can be solved and the effect can be obtained even if some components are deleted from all the components shown in the embodiment, the configuration from which the components are deleted can be extracted as an invention. [Explanation of symbols]

[0093] 1...Model learning device 2...External device 10...Input data processing section 20...Setting parameter processing section 30…Model parameter learning section 40...Model parameter processing unit 50…Recording section 51...Data recording section 52...Setting parameter recording section 53...Model parameter recording section 60...Input / output section 110... Input / Output Interface 120…CPU 121...Control device 122...Arithmetic device 130...Storage device 131...Main memory 132…Auxiliary storage device 140…Bus

Claims

1. an input data processing unit for acquiring input data; a configuration parameter processing unit that acquires configuration parameters; a model parameter learning unit that uses the input data and the setting parameters as inputs, and rank-learns model parameters using an objective function that is composed of a loss function that maintains an order relationship and a regularization term that limits an output of a model; A model parameter processing unit that outputs learned model parameters; The input data is observable multivariate data and label data indicating an order relationship; The setting parameters include hyperparameters of the regularization terms, The model parameter learning unit obtains model parameters that output a score for predicting a dependent variable. Model learning device.

2. The hyperparameter includes a quantile of the objective variable, where the quantile is a number that exists at a position where a set of sorted numbers is equally divided by positive integers; the model parameter learning unit approximates a quantile of an output value of a model to the quantile of the response variable; The model learning device according to claim 1 .

3. the model parameter learning unit minimizes an objective function including a term calculated using the quantiles of the objective variable and the quantiles of the output value of the model. The model learning device according to claim 2 .

4. The hyperparameters further include a parameter that determines a weight of the regularization term. The model learning device according to claim 2 or 3.

5. The model parameter learning unit optimizes the objective function using a gradient method, a stochastic gradient method, or Adam.

5. The model learning device according to claim 1.

6. The setting parameters further include a learning rate parameter used in optimizing the objective function. The model learning device according to claim 5 .

7. Obtaining input data; Obtaining configuration parameters; rank-learning model parameters using an objective function including a loss function for maintaining an order relationship and a regularization term for limiting an output of a model, with the input data and the setting parameters as input; Outputting the learned model parameters; The input data is observable multivariate data and label data indicating an order relationship; The setting parameters include hyperparameters of the regularization terms, The rank learning obtains model parameters that output scores predicting the dependent variable. A computer implemented method for learning models.

8. A model learning program for causing a computer to execute the functions of each component of the model learning device according to any one of claims 1 to 6.

Images