Research On Stock Index Prediction Based On Bayesian-Gray Forecasting Theory

As China's economic strength gradually improves,the people's living standards are increasing,and people's pockets are getting more and more drums.Now that more and more people have started to learn financial management,the most attention will be given to the stock market.The importance of the stock market for a country is beyond doubt.The development history of China's finance can basically be used as a microcosm through the development of the stock market.There are more and more listed companies in the stock market,more and more funds in stock investment,and more and more people are concerned about the stock market,which means that more and more people are studying the stock market.The development of the financial market affects the hearts of all investors,and it also affects the heart of the government regulators.In 2015,China's "stock-related disasters" still vividly reflected,and the stock market is indeed one of the best places for financial investment.However,its high-risk nature cannot be ignored by us.So how do we research the stock market and get the financial industry and academics?The wide attention of the community.This article through the observation of the daily status of the stock market,from a macro perspective and a micro perspective analysis of the operation of the stock market,put forward in a certain period of time,the stock market will be in a closed state of view,which innovatively proposed The Bayesian method is used to optimize the forecast window length and then combined with the gray model to construct the predictive model.This paper describes in detail the principles of the GM(1,1),gray Verhulst model,and discrete grey Verhulst model,as well as their characteristics.After conducting preliminary empirical studies using Bayes-GM(1,1),Bayesian-Grey Verhulst model,Bayesian-discrete gray Verhulst model,and three single gray models,we found that when we use gray Verhulst model to forecast the Shanghai Composite Index,there will be a situation where the model's prediction results deviate seriously from the true value at a certain point in time.For this reason,the specific reasons for the serious deviation of the model's predicted value are analyzed,and the Bayesian and mixed grey models are proposed boldly.The combined method constructs Bayesian-GM(1,1)-Verhulst prediction model,Bayes-GM(1,1)-discrete Verhulst prediction model,and hybrid prediction model without Bayesian.Through empirical analysis,it is proved that the hybrid model has better prediction ability.Whether it is preliminary empirical analysis or empirical analysis of the latter optimization model,the results of the Bayesian method generated by the gray prediction window length are simply verified that the Shanghai Composite Index will show consistency at a certain time.