Abstract: According to the characteristics of the stock index futures prediction, the indicators that have great influence on the development trend of stock index futures are selected and the support vector machines are firstly used to the regression prediction of stock index futures. Besides, genetic algorithm (GA) and particle swarm optimization algorithm (PSO) are employed to optimize the support vector machine (SVM) with four different kernel functions and eight different programs are attained. By comparing the accuracy and the time complexity of all the programs, the empirical study shows that the linear kernel function SVM optimized by PSO is the best model for regression prediction of Chinese stock index futures.

Key words: Chinese stock index futures, support vector machine (SVM), genetic algorithm (GA), particle swarm optimization algorithm (PSO), regression prediction