Fuzzy ARTMAP with Feature Weighting

R. Andonie (USA), A. Caţaron, and L.M. Sasu (Romania)


Fuzzy ARTMAP, feature weighting, classification, ma chine learning


We introduce a novel Fuzzy ARTMAP (FAM) architec ture: FAM with Feature Weighting (FAMFW). In the first stage, the features of the training data are weighted. In the second stage, the obtained weights are used to im prove the FAMFW training. The effect of this approach is a more sensitive FAM category determination: Cate gory dimensions in the direction of relevant features are decreased whereas category dimensions in the direction of non-relevant feature are increased. Potentially, any fea ture weighting method could be used, which makes the FAMFW very general. In our study, we use a feature weighting algorithm based on the Neural-Gas algorithm.

Important Links:

Go Back