PROCEEDINGS IPMU '08
A new fuzzy 3-rules pattern classifier with reject options based on aggregation of membership degrees
Hoel Le Capitaine, Carl Frélicot
In this paper, we address the problem
of fuzzy rule-based pattern recognition
with reject options. These options are
made possible thanks to simple rules
whose satisfaction level is expressed by
the value of dedicated operators that aggregate
degrees of typicality. Results
obtained with the proposed classifier on
artificial and real data are given.
 | PDF full paper |