We present an approach for the supervised online learning of object representations based on a biologically motivated architecture of visual processing. We use the output of a rece...
Abstract. This paper presents a handwritten Greek character recognizer. The recognizer is composed of two modules: the first one is a feature extractor, the second one, the classi...
This paper introduces multiple instance regression, a variant of multiple regression in which each data point may be described by more than one vector of values for the independen...
Various alternatives have been developed to improve the Winner-Takes-All (WTA) mechanism in vector quantization, including the Neural Gas (NG). However, the behavior of these algo...
Aree Witoelar, Michael Biehl, Anarta Ghosh, Barbar...
This paper presents an unsupervised fuzzy-kernel learning vector quantization algorithm called FKLVQ. FKLVQ is a batch type of clustering learning network by fusing the batch learn...