Abstract: With the traditional SVMs' two-class problems extending to the multi-class problems, the unclassifiable regions exist. Flight action recognition is proposed in this paper based on the fuzzy support vector Machines(FSVMs). Six kinds of recorded flight data of flight action are recognized by FSVMs. Simulation results show that the improved FSVMs has characteristics of simpler calculation and better testing correctness than the traditional SVMs approach .

Key words: support vector machines, fuzzy support vector machines, flight action recognition