基于支持向量机的上证指数开盘指数预测

支持向量机在模式分类、回归分析等领域应用的过程中，其性能依赖于参数的选取，而参数往往是随机给定的或凭测试经验给定。针对这一情况，对影响模型分类、回归能力的相关参数进行了研究，采用交叉验证的思想对其中重要的两个参数（误差惩罚参数和高斯核参数）进行了优化，并成功的应用于上证指数开盘指数的预测，达到了很好的效果。

Prediction of Opened Index in Shanghai Composite Index Based on Support Vector Machines

In the application of pattern classification,regression analysis and other fields,the performance of Support Vector Machine always depends on the parameter selection,and often given by random or experience.In response to this situation,the relevant parameters which affect the model of classification,regression ability had been studied,using Cross-Validation,two of the best important parameters (error penalty parameter and Gaussian kernel parameter) is optimized,which can successfully used in the prediction ...