粉体の僅かな変化を機械学習（mt法）で検知する

Abstract

In order to functionalize the powders, it is important to understand the relationship between the powders' structure and their functionalities. As one of the techniques, the Mahalanobis-Taguchi system (MTS) has been utilized because that effective variables which improve accuracy of classification can be visualized. The surface-roughness-related variables were the effective powder structure which improve accuracy of classification into male/female of Japanese beetles larval droppings as the model powders. Considering that binder concentration in the females' dropping was higher than in the males' dropping, the females' droppings could be deformed during molding process through the larval gut. The MTS can be expected as the effective way to improve the accuracy of the classification by visualization of the relationship between powder structure and functionalities without individual differences.