Vorträge

Data Quality Measures in Decision-Making - A Fuzzy ARTMAP Approach

Prof. Razvan Andonie
Department of Electronics and Computers,
Transylvania University, Brasov, Romania

Abstract:

An incremental, nonparametric probability estimation procedure using a variation of the fuzzy ARTMAP neural network is introduced. The resulted network, called FAMr, uses a relevance factor assigned to each sample pair, proportional to the importance of the respective pair in learning. We prove that our incremental probability estimator is correct. Experimental results have shown that FAMr compares favorably with FAM and PFAM, both as a classifier and a probability estimator. The relevance factor can be used as a noise filter, attenuating signals from a nonreliable source. The performance of the network depends on the choice of certain relevance factor values.

Zeit: Mittwoch, den 13. November 2002, 17 Uhr c.t.

Ort: Institut für Neuro- und Bioinformatik
        TZL, Seelandstr. 1a, Geb. 5
        Seminarraum OG.3