Title :
An improved BP algorithm for pattern recognition
Author :
Yujuan, Zeng ; Xiangdong, Wang ; Shoujue, Wang
Author_Institution :
Inst. of Semicond., Acad. Sinica, Beijing, China
Abstract :
An improved backpropagation (BP) algorithm for pattern recognition is proposed in this paper. Using a function substitution for an error measure, it resolves the inconsistency of the 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 trebled and the performance of the network is better than the conventional BP algorithm with momentum and adaptive step size
Keywords :
backpropagation; convergence; neural nets; pattern recognition; BP algorithm; backpropagation algorithm; convergence speed; error measure; function substitution; pattern recognition; quadratic error; training pattern; Computer errors; Computer simulation; Content addressable storage; Convergence; Error correction; Feedforward neural networks; Feedforward systems; Neural networks; Optimization methods; Pattern recognition;
Conference_Titel :
Signal Processing Proceedings, 1998. ICSP '98. 1998 Fourth International Conference on
Conference_Location :
Beijing
Print_ISBN :
0-7803-4325-5
DOI :
10.1109/ICOSP.1998.770864
