Based On The Evolution Of Causal Elastic Energy Stocks Form Ridge Regression Prediction Research

As the stock market volatility respectively, the stock market situation in effective prediction is still a problem. Due to the stock volatility has certain shape characteristic, these form can effectively contain the stock market respectively, ascending the trend prediction effect; Paper base on morphological characteristics in the stock market, combines the technique of causal analysis and evolution of elastic theory, and giving the idea of complex systems from the energy and space to study the stock market trend prediction. Project of research work has high display significance and application value. The specific research work are as follows:(1) Traditional stock market forecasting method is difficult to effectively forecast the stock market situation, therefore, based on the typical form of M on stock, put forward a kind of ridge regression predict stock market momentum algorithm based on causality algorithm (RRC). RRC algorithm basic idea:according to the characteristics of the fluctuations in the form of M,combined with energy thought,then according M form’s peaks and troughs for node, build a Bayesian network structure model of M form; Then, by using Markov blanket algorithm and asymmetric information entropy,learn local causal structure of M form;, thus, introduction the causal strength metrics, and introduce the M shape causality in ridge regression models to predict stock market trend, and the stock market trend prediction algorithm is given.(2) During the formation of M form frequently display volatility, platform, lianyang,lianyin, etc, the node and edge energy conversion direction protean. Aiming at this defect, on the basis of previous studies, based on the elastic energy evolution ridge regression stock market trend prediction algorithm (EE-RR) is proposed. The algorithm of the basic idea is:first build a Bayesian network structure in the form of M, looking for a target variable Markov blanket; Then, combined with the energy of elastic evolution, build M form the energy of the nodes elastic evolution model; Finally will M form vertex and edge of the elastic energy evolution model combined with ridge regression model, the stock market M shape prediction algorithm is given.On the Shangzheng index and the Shenzhen index selected respectively on the experimental data of the above two algorithms, this paper compares and analyzes the experimental results show that the above two algorithms in the stock market on the forecast effect is better than the average of prediction algorithm, EE-RR algorithm for predicting effect in the formation of M is better than RRC algorithm.