Applications of simulated annealing-immune particle swarm optimization in emotion recognition of galvanic skin response signal
ZHOU Yu-ting1, LIU Guang-yuan1, LAI Xiang-wei2
1.School of Electronics and Information Engineering, Southwest University, Chongqing 400715, China
2.School of Computer and Information Science, Southwest University, Chongqing 400715, China
Abstract��An improved immune particle swarm optimization was presented in this study in order to increase the effectiveness of feature selection for emotion recognition based on Galvanic Skin Response (GSR). Firstly, 342 groups of GSR signal with each containing 6 kinds of affective data were denoised, and afterwards the original features were extracted. Then simulated annealing mechanism was introduced to the particle update process of Immune Particle Swarm Optimization (IPSO), and the Simulated Annealing-Immune Particle Swarm Optimization (SA-IPSO) was adopted for feature selection. The experimental results show that compared with IPSO, SA-IPSO can achieve relatively high recognition rate with fewer features, the application of simulated annealing mechanism can help with the optimization of feature selection, and the improved algorithm also performs well on global convergence.