Paper 70: "Class-to-class Siamese Networks: Explaining Classification

19 views /


Case-based classification methods normally assess the similarity between a query and class members in the case base. Such methods can exploit data-driven machine learning methods, such as siamese networks, to learn to detect similarity. This paper proposes a difference-based approach, {\it class-to-class siamese network} (C2C-SN),...


0 commentaries