作者: Th. Villmann , B. Hammer , F. -M. Schleif , T. Geweniger , T. Fischer
DOI: 10.1007/11893257_5
关键词: Vector quantization 、 Computer science 、 Generalization error 、 Stability (learning theory) 、 Linear classifier 、 Supervised learning 、 Wake-sleep algorithm 、 Mutual information 、 Learning vector quantization 、 Competitive learning 、 Artificial neural network 、 Semi-supervised learning 、 Unsupervised learning 、 Artificial intelligence 、 Fuzzy logic 、 Fuzzy classification 、 Machine learning
摘要: In this article we extend the (recently published) unsupervised information theoretic vector quantization approach based on Cauchy–Schwarz-divergence for matching data and prototype densities to supervised learning classification. particular, first generalize method more general metrics instead of Euclidean, as it was used in original algorithm. Thereafter, model a resulting fuzzy classification Thereby, allow labels both, prototypes. Finally, transfer idea relevance metric adaptation known from new approach.