沪深300指数马氏性检验及预测

利用粗粒化方法构造沪深300指数和交易量的联合变动状态序列,通过计算χ2统计量检验了序列的马氏性,选取标准化的各阶收益率自相关系数作为权重,建立加权马氏链预测模型并对股指走势进行预测。结果发现成交量在预测中具有重要作用,日内走势预测效果较好,日间走势预测效果较差,说明股指日间走势受市场消息面影响较大。

Prediction and analysis of Shanghai and Shenzhen 300 index based on weighted Markov chain

Using coarse-grained method to construct the joint movements sequence of CSI 300 Index and volume,through a chi-square test verification sequence is a markov chain and select standardization various order yields since correlation coefficients as weights.Then establishing the weighted Markov chain model to forecasting stock index trend.The results show that volume plays an important role in predicting stock trend, daytime better effects and days trend prediction effect is poorer the model of weighted Markov chain is more effective in short-term forecasts predict than long.