Storedproducted Insects's Voice Signals Classification Based on the Chaos Optimization

In order to improve the automatic recognition rate of voice signals with respect to stored product insects,this paper proposed support vector machine model parameters automatic selection algorithm based on chaotic data series to search for the optimal combination of support vector machine model parameters C and RBF kernel width parameters σ.The parameters(C,σ) have a great impact on the generalization ability.Based on the chaotic sequence generated by Logistic Map and Circle Map,respectively,the chaotic variables are zoomed to the parameter(C,σ) value space to search for the optimal combination.Compared with the grid method,experimental results show that this method can improve the classification rate,and significantly reduce the number of training support vector machines,so the support vector machine has better generalization ability.