Optimization apparatus, non-transitory computer-readable storage medium for storing optimization program, and optimization method
A method includes: partitioning learning data containing objective variables and explanatory variables into a plurality of subsets of data; executing regularization processing on first data in each of the partitioned subsets, and extracting a first element equal to zero; extracting, as a candidate, each model where an error ratio between first multiple regression and second multiple regression is equal to or more than a predetermined value, the first multiple regression being a result of multiple regression on second data which is test data in each of the partitioned subsets and is for use to calculate the error ratio of the learning data, the second multiple regression being a result of multiple regression on third data obtained by excluding the first element from the second data; and outputting a model where zero is substituted for an element that takes zero a predetermined or larger number of times in the candidate.
1. An optimization apparatus comprising:
a memory; and
a processor coupled to the memory, the processor being configured to perform processing including:
partitioning learning data containing objective variables and explanatory variables into a plurality of subsets of data;
executing regularization processing on first data to be used for structure extraction of the learning data in each of the partitioned subsets of data, and extracting a first element equal to zero;
extracting, as a candidate model, each model in which an error ratio between first multiple regression and second multiple regression is equal to or more than a predetermined value, the first multiple regression being a result of multiple regression on second data which is test data in each of the partitioned subsets of data and is for use to calculate the error ratio of the learning data, the second multiple regression being a result of multiple regression on third data obtained by excluding the first element from the second data; and
outputting a model in which zero is substituted for an element that takes zero a predetermined or larger number of times in the extracted candidate models.
2. The optimization apparatus according to claim 1 , wherein the executing of regularization processing is configured to perform the regularization processing by annealing data expressed in an ising format based on the first data.
3. The optimization apparatus according to claim 2 , the processing further comprising setting, when l denotes a sequence length to be used in binary expansion for expression in the ising format, Nb denotes an upper limit number of bits in the annealing, and n denotes the number of rows in the learning data, the sequence length to be used for the binary expansion to l that is the largest integer satisfying n(l+2)≤Nb.
4. The optimization apparatus according to claim 1 , wherein the partitioning is configured to partition the learning data into the subsets of data such that an upper limit and a lower limit of coefficients contained in each of the subsets of data after the learning data partitioning satisfy a predetermined condition.
5. An optimization method comprising:
partitioning learning data containing objective variables and explanatory variables into a plurality of subsets of data;
executing regularization processing on first data to be used for structure extraction of the learning data in each of the partitioned subsets of data, and extracting a first element equal to zero;
extracting, as a candidate model, each model in which an error ratio between first multiple regression and second multiple regression is equal to or more than a predetermined value, the first multiple regression being a result of multiple regression on second data which is test data in each of the partitioned subsets of data and is for use to calculate the error ratio of the learning data, the second multiple regression being a result of multiple regression on third data obtained by excluding the first element from the second data; and
outputting a model in which zero is substituted for an element that takes zero a predetermined or larger number of times in the extracted candidate models.
6. A non-transitory computer-readable storage medium for storing an optimization program which causes a processor to perform processing for object recognition, the processing comprising:
partitioning learning data containing objective variables and explanatory variables into a plurality of subsets of data;
executing regularization processing on first data to be used for structure extraction of the learning data in each of the partitioned subsets of data, and extracting a first element equal to zero;
extracting, as a candidate model, each model in which an error ratio between first multiple regression and second multiple regression is equal to or more than a predetermined value, the first multiple regression being a result of multiple regression on second data which is test data in each of the partitioned subsets of data and is for use to calculate the error ratio of the learning data, the second multiple regression being a result of multiple regression on third data obtained by excluding the first element from the second data; and
outputting a model in which zero is substituted for an element that takes zero a predetermined or larger number of times in the extracted candidate models.