FEATURE QUANTITY SELECTING METHOD, FEATURE QUANTITY SELECTING PROGRAM, FEATURE QUANTITY SELECTING DEVICE, MULTICLASS CLASSIFICATION METHOD, MULTICLASS CLASSIFICATION PROGRAM, MULTICLASS CLASSIFICATION DEVICE, AND FEATURE QUANTITY SET

摘要

The objective of the present invention is to provide a multiclass classification method, a multiclass classification program, and a multiclass classification device capable of classifying a sample having a plurality of feature quantities into any of a plurality of classes, robustly and with a high degree of accuracy, on the basis of the value of a selected portion of the feature quantities. Another objective of the present invention is to provide a feature quantity selecting method, a feature quantity selecting program, a feature quantity selecting device, and a feature quantity set used for such multiclass classification. The present invention deals with the problem of multiclass classification involving feature quantity selection. Feature quantity selection is literally a method for selecting, in advance, feature quantities necessary for each subsequent process (in the present invention, multiclass classification in particular), from among multiple feature quantities included in a sample. Multiclass classification is a discrimination problem for determining whether a given unknown sample belongs to any of a plurality of classes.