Study Shanghai stock index forecasting problem. The stock index is stochastic, time-varying, and with bigger fluctuation, therefore, the forecasting accuracy of traditional linear method is very low. This paper put forward a stock index forecasting method based on support vector machine. The support vector machine based on statistical theory and structural risk minimization principle can overcome the shortcomings of neural network algorithm, such as over fitting and local optimal defects, and has excellent generalization ability. The performance of the proposed algorithm was tested with Shanghai Composite Index from 1990 to 2009. The simulation results show that the support vector machine has higher precision, smaller error in stock index prediction, and the forecasting results can provide valuable advice for users.%研究上证指数预测问题,针对证券指数变化具有随机性、时变、波动性较大等特点,传统线性预测方法预测精度低等缺陷,提出一种基于支持向量机的证券指数预测方法.支持向量机是一种基于统计学理论和结构风险最小化原则的机器学习方法,克服了类似神经网络经验见险最小化原则算法的过拟合、局部极值等缺陷,泛化能力优异.采用1990 ～ 2009年上证指数对算法性能进行测试,仿真结果表明,支持向量机是一种预测精度高、误差小的证券指数预测算法,预测结果可以为用户提供有价值的参考意见.
展开▼
