Pu classifier, Pu classification method and Pu classification program

摘要

Provision of PU classification device, PU classification method, and PU classification program. Given a case to be classified, the magnitude relationship between the first probability that the case is sampled from the learning population distribution as a positive case and the second probability that the case is sampled from the learning population distribution. The distribution function of the first probability from the classifier that classifies the cases to be classified into positive cases or negative cases most likely using the judgment inequality, and the set of positive cases sampled from the learning population distribution. Is provided, and a learning unit that learns the classifier by estimating the distribution function of the second probability from a set of cases whose positive and negative values are unknown sampled from the learning population distribution is provided, and learning is performed by the learning unit. The cases to be classified are classified into positive cases or negative cases by using the above-mentioned classifier.