Relevance determination in learning vector quantization
Thorsten Bojer, Barbara Hammer, Daniel Schunk, Katharina Tluk von Toschanowitz · 2001
Abstract. We propose a method to automatically determine the rel-evance of the input dimensions of a learning vector quantization (LVQ) architecture during training. The method is based on Hebbian learning and introduces weighting factors of the input dimensions which are auto-matically adapted to the specic problem. The benets are twofold: On the one hand, the incorporation of relevance factors in the LVQ archi-tecture increases the overall performance of the classication and adapts the metric to the specic data used for training. On the other hand, the method induces a pruning algorithm, i.e. an automatic detection of the input dimensions which do not contribute to the overall classier. Hence we obtain a possibly more eÆcient classication and we gain insight to the role of the data dimensions. 1.