MACHINE LEARNING APPARATUS, MACHINE LEARNING METHOD, CLASSIFICATION APPARATUS, CLASSIFICATION METHOD, AND PROGRAM

摘要

PROBLEM TO BE SOLVED: To classify contents to belonging categories accurately.SOLUTION: An image acquisition section 121 of a machine learning apparatus 100 acquires n (n is a natural number equal to or larger than 2) labeled learning images to be used for categorization. A feature vector acquisition section 122 acquires a feature vector indicating characteristics, from each of the n learning images. A vector conversion section 123 converts a feature vector of each of the n learning images to a similar feature vector, on the basis of similarity between learning images. A classification condition learning section 125 learns a classification condition for classifying the n learning images by category, on the basis of the similar feature vector converted by the vector conversion section 123 and a label added to each of the n learning images. A classifying section 126 classifies unlabeled test images by category, according to the classification condition learned by the classification condition learning section 125.SELECTED DRAWING: Figure 1