Feature dimension reduction based on genetic algorithm for mandarin digit recognition

The dimensions are higher after combining Mel frequence cepstral coefficient with linear prediction cepstrum coefficient. In this paper, genetic algorithm is proposed to reduce the dimensions of the feature data to improve recognition performance of the system. First, extract Mel frequence cepstral coefficient and linear prediction cepstrum coefficient of the speech signal; then, reduce the dimensions of the feature data based on genetic algorithm; finally, the low dimensional data are sent into the support vector machine. Simulation results demonstrate that the recognition rate increases by 12.2% using genetic algorithm compared with principle component analysis, recognition rate almost has no change compared with the initial characteristics and the recognition speed gets improved effectively.