An improved BP algorithm for pattern recognition

An improved BP algorithm for pattern recognition is proposed in this paper. By a function substitution for error measure, it resolves the inconsistency of BP algorithm for pattern recognition problems, i.e. the quadratic error is not sensitive to whether the training pattern is recognized correctly or not. Trained by this new method, the computer simulation result shows that the convergence speed is increased to treble and performance of the network is better than conventional BP algorithm with momentum and adaptive step size.

An improved BP algorithm for pattern recognition is proposed in this paper. By a function substitution for error measure, it resolves the inconsistency of BP algorithm for pattern recognition problems, i.e. the quadratic error is not sensitive to whether the training pattern is recognized correctly or not. Trained by this new method, the computer simulation result shows that the convergence speed is increased to treble and performance of the network is better than conventional BP algorithm with momentum and adaptive step size.

于2010-10-29批量导入

Made available in DSpace on 2010-10-29T06:37:12Z (GMT). No. of bitstreams: 1 2997.pdf: 253233 bytes, checksum: d908859e12567303f2f3d201c394ef3e (MD5) Previous issue date: 1998

Chinese Inst Electr.; Signal Proc Soc.; IEEE Signal Proc Soc.; Inst Elect Engineers.; Union Radio Sci Int.; IEEE Beijing Sect.; Natl Nat Sci Fdn China.; CIE Comm URSI.; IEEE Comp Soc Beijing Chapter.; IEEE SP Soc Beijing Chapter.

Chinese Acad Sci, Inst Semicond, Beijing 100083, Peoples R China

Chinese Inst Electr.; Signal Proc Soc.; IEEE Signal Proc Soc.; Inst Elect Engineers.; Union Radio Sci Int.; IEEE Beijing Sect.; Natl Nat Sci Fdn China.; CIE Comm URSI.; IEEE Comp Soc Beijing Chapter.; IEEE SP Soc Beijing Chapter.