MISSING VALUES AND LEARNING OF FUZZY RULES

In this paper a technique is proposed to tolerate missing values based on a system of fuzzy rules for classification. The presented method i s mathematically solid but nevertheless easy and efficient to implemen t. Three possible applications of this methodology are outlined: the c lassification of patterns with an incomplete feature vector, the compl etion of the input vector when a certain class is desired, and the tra ining or automatic construction of a fuzzy rule set based on incomplet e training data. In contrast to a static replacement of the missing va lues, here the evolving model is used to predict the most possible val ues for the missing attributes. Benchmark datasets are used to demonst rate the capability of the presented approach in a fuzzy learning envi ronment.