ELLIPTIC BASIS FUNCTION PROBABILISTIC NEU RAL-NETWORK FOR CHARACTER-RECOGNITION

An Elliptic Basis Function Probabilistic Neural Network (EBPNN) model for character recognition with noise was proposed, which uses selectiv e attentional parameters extraced from statistic features of character s as elliptic basis function parameters. EBPNN is not only used for bi nary pattern recognition, but also for continuous pattern recognition. The experiments show that the recognition rate of EBPNN is better tha n that of other networks, especially under the noise circumstances.