Extraction of Classification Patterns from Deep Learning Networks

Abstract

In deep learning, there is a problem that concrete classification patterns for deriving reasons for classification are often incomprehensible. In this paper, we propose a classification patterns extraction system from deep learning networks and verified the effectiveness of the system. The proposed system takes out learning networks from the learning result of deep learning and extracts classification patterns from the learning networks. Then the system displays the extracted classification patterns so that users of the system can interpret the learning networks. In verification experiments, the significance of the extracted classification patterns was estimated by chi-square test. The results showed that users of the system can extract classification patterns effective for interpretations of the learning networks by using the proposed system.