Speech Recognition Based on Fuzzy Neural Network and Chaotic Differential Evolution Algorithm

摘要

The training method is very important in speech recognition. But the traditional BP neural network method trained for a long time, easily falling into local extreme disadvantage. However, the optimization algorithm can significant improve the model performance like response speed and recognition accuracy. In this paper, we propose a Chaotic Differential Evolution (CDE) algorithm for optimization the objective function of the fuzzy neural network model in training process, which can effectively improve the accuracy and consistency of speech recognition. The results show that the CDE algorithm optimized fuzzy neural network has more convergence speed and recognition rate than PSO and BP algorithm.