意愿计算的股市突变点预测方法

由于传统分段线性表示方法没有考虑股市数据分布变化导致分段不合理,同时股市突变点相关特征的局部性导致突变点难以有效预测,所以在分段线性表示方法的基础上提出一种意愿计算的股市突变点预测方法(WC-WSVM)。首先,给出一种波动率分布变化的分段线性表示(V-PLR)方法,通过波动率分布变化自适应地优化PLR分段阈值;然后,提取与主力买卖股票意愿相关的股市特征并进行量化,利用逻辑回归(LR)对于所提取的特征进行融合得到意愿计算结果;最后,将意愿计算结果与PLR-WSVM算法输入特征共同代入到WSVM中,进行突变点预测。在真实数据上的实验结果表明,算法具有强适应性,预测精度得到有效提升。

Prediction method of stock market mutation point prediction with willingness calculation

Because the traditional piecewise linear representation method does not consider the stock market data distribution changes,which leads to unreasonable segmentation.Meanwhile the locality of the stock market mutation point related characte-ristics makes the mutation point difficult to predict effectively.Therefore,based on the piecewise linear representation method,this paper proposed a prediction method of stock market mutation point with willingness calculation(WC-WSVM).Firstly,it gave a piecewise linear representation(V-PLR)method of volatility distribution change,and the volatility distribution change optimized the PLR segmentation threshold.Secondly,itextracted the stock market characteristics related to the main willingness to buy and sell stocks to quantify,using logistic regression(LR)to fuse the extracted features to obtain the willingness calculation results.Finally,it brought the willingness calculation results and the PLR-WSVM algorithm input features into the WSVM together,to predict the mutation point.Experimental results on real data show that the algorithm is more adaptable and the prediction accuracy is effectively improved.