Information processing apparatus, program, and information processing method

1. A method for training a prediction model for predicting a label to be allocated to a prediction target explanatory variable set, the method comprising:

acquiring a set of training data comprising a set of explanatory variables and a label allocated to the set of explanatory variables;

extracting a plurality of explanatory variable subsets from the set of explanatory variables, wherein respective ones of the plurality of explanatory variable subsets are allocated a different weight coefficient;

generating a feature vector, concerning each of the plurality of explanatory variable subsets, the feature vector comprising a plurality of feature values;

executing Bayesian inference using prior distributions of a regression vector comprising a plurality of regression coefficients respectively corresponding to the plurality of feature values of the feature vector;

training a prediction model utilizing the plurality of explanatory variable subsets, the different weight coefficients, the feature vector, the regression vector, and an objective function to be reduced, the objective function comprising a weighted sum of terms indicating errors between labels predicted for the plurality of explanatory variable subsets based on the feature vector and the regression vector, and the label allocated to the set of explanatory variables; and

outputting a posterior probability distribution of the regression vector and the different weight coefficients as a training result.

2. The method according to claim 1 , wherein training the prediction model further comprises:

utilizing, as prior distributions, the posterior probability distribution of the regression vector and the weight coefficients; and

outputting posterior probability distributions of the regression vector and the weight coefficients for additional training data.

3. The method according to claim 1 , wherein the set of training data is a time-series data set observed over time, and wherein the extracting further comprises:

extracting, as the plurality of explanatory variable subsets, a plurality of data sequences continuous in time series.

4. The method according to claim 3 , wherein the plurality of data sequences comprises a set of values of a plurality of feature values in a plurality of sections.

5. The method according to claim 3 , wherein the plurality of data sequences partially overlap one another in a time series.

6. The method according to claim 1 , wherein the acquiring further comprises:

acquiring a prediction target data set comprising the prediction target explanatory variable set, and wherein the method further comprises:

generating the predicted label corresponding to the prediction target explanatory variable set based on the prediction model.

7. The method according to claim 6 , wherein the training further comprises:

setting, as additional training data, the prediction target data set; and

further training the prediction model based on the prediction target data set.

8. An information processing apparatus for training a prediction model for predicting a label to be allocated to a prediction target explanatory variable set, the information processing apparatus comprising:

a memory;

a processor communicatively coupled to the memory, wherein the processor is configured to perform operations of:

acquire a set of training data, the set of training data comprising a set of explanatory variables and a label allocated to the set of explanatory variables;

extract a plurality of explanatory variable subsets from the set of explanatory variables, wherein respective ones of the plurality of explanatory variable subsets are allocated a different weight coefficient;

generate a feature vector, concerning each of the plurality of explanatory variable subsets, the feature vector comprising a plurality of feature values;

executing Bayesian inference using prior distributions of a regression vector comprising a plurality of regression coefficients respectively corresponding to the plurality of feature values of the feature vector;

train a prediction model utilizing the plurality of explanatory variable subsets, the different weight coefficients, the feature vector, the regression vector, and an objective function to be reduced, the objective function comprising a weighted sum of terms indicating errors between labels predicted for the plurality of explanatory variable subsets based on the feature vector and the regression vector, and the label allocated to the set of explanatory variables; and

output a posterior probability distribution of the regression vector and the different weight coefficients as a training result.

9. The information processing apparatus according to claim 8 , wherein the processor configured to acquire is further configured to:

acquire a prediction target data set comprising the prediction target explanatory variable set, and wherein the information processing apparatus is further configured to:

generate the predicted label corresponding to the prediction target explanatory variable set based on the prediction model.

10. The information processing apparatus according to claim 8 , wherein the processor configured to train the prediction model is further configured to:

utilize, as prior distributions, the posterior probability distribution of the regression vector and the weight coefficients; and

output posterior probability distributions of the regression vector and the weight coefficients for additional training data.

11. The information processing apparatus according to claim 8 , wherein the set of training data is a time-series data set observed over time, and wherein the processor configured to extract is further configured to:

extract, as the plurality of explanatory variable subsets, a plurality of data sequences continuous in time series.

12. The information processing apparatus according to claim 11 , wherein the plurality of data sequences comprises a set of values of a plurality of feature values in a plurality of sections.

13. The information processing apparatus according to claim 11 , wherein the plurality of data sequences partially overlap one another in a time series.

14. The information processing apparatus according to claim 9 , wherein the processor configured to train is further configured to:

set, as additional training data, the prediction target data set; and

further train the prediction model based on the prediction target data set.

15. A computer program product for training a prediction model for predicting a label to be allocated to a prediction target explanatory variable set, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a computer to cause the computer to perform a method comprising:

acquiring a set of training data comprising a set of explanatory variables and a label allocated to the set of explanatory variables;

extracting a plurality of explanatory variable subsets from the set of explanatory variables, wherein respective ones of the plurality of explanatory variable subsets are allocated a different weight coefficient;

generating a feature vector, concerning each of the plurality of explanatory variable subsets, the feature vector comprising a plurality of feature values;

executing Bayesian inference using prior distributions of a regression vector comprising a plurality of regression coefficients respectively corresponding to the plurality of feature values of the feature vector;

training a prediction model utilizing the plurality of explanatory variable subsets, the different weight coefficients, the feature vector, the regression vector, and an objective function to be reduced, the objective function comprising a weighted sum of terms indicating errors between labels predicted for the plurality of explanatory variable subsets based on the feature vector and the regression vector, and the label allocated to the set of explanatory variables the set of training data, wherein the prediction model generates a predicted label to be allocated to a prediction target explanatory variable set with each of the plurality of explanatory variable subsets allocated a different weight coefficient and a regression vector comprising a plurality of regression coefficients respectively corresponding to the plurality of feature values of the feature vector and the different weight coefficient of each of the plurality of explanatory variables; and

outputting a posterior probability distribution of the regression vector and the different weight coefficients as a training result.

16. The computer program product according to claim 15 , wherein the program instructions configured for training the prediction model comprise additional program instructions executable by the computer to cause the computer to perform the method further comprising:

utilizing, as prior distributions, the posterior probability distribution of the regression vector and the weight coefficients; and

outputting posterior probability distributions of the regression vector and the weight coefficients for additional training data.

17. The computer program product according to claim 15 , wherein the set of training data is a time-series data set observed over time, and wherein the program instructions for extracting comprises additional program instructions executable by the computer to cause the computer to perform the method further comprising:

extracting, as the plurality of explanatory variable subsets, a plurality of data sequences continuous in time series.

18. The computer program product according to claim 17 , wherein the plurality of data sequences comprises a set of values of a plurality of feature values in a plurality of sections.

19. The computer program product according to claim 17 , wherein the plurality of data sequences partially overlap one another in a time series.

20. The computer program product according to claim 15 , wherein the program instructions for acquiring comprise additional program instructions executable by the computer to cause the computer to perform the method further comprising:

acquiring a prediction target data set comprising the prediction target explanatory variable set, and wherein the method further comprises:

generating the predicted label corresponding to the prediction target explanatory variable set based on the prediction model; and

wherein the program instructions for training comprise additional program instructions executable by the computer to cause the computer to perform the method further comprising:

setting, as additional training data, the prediction target data set; and

further training the prediction model based on the prediction target data set.