Stock Index Prediction Based on the Analytical Center of Version Space
基于可行域解析中心的股指时间序列预测

We attempt to transform predicting the value into predicting the range of stock index value, which yields a multi-class classification problem. Then we review multi-class classifiers commonly adopted and propose a multi-class classifier based on analytical center of version space. Experiments on shanghai stock index demonstrates that the multi-class classifier is validated and the method proposed of stock index prediction of practical value.