Research And Application Of Key Technologies Based On Hidden Markov Prediction

With the development of modern society and the gradual improvement of science and technology,daily material life has been greatly enriched.People are no longer satisfied with the simple material needs of food,clothing,shelter and transportation,and gradually turn the surplus wealth from savings to investment.Stock investment becomes the first choice of contemporary people.But investment income and risk tend to change direction,so how to ensure that the return on investment under the condition of reasonable reducing risk will be a part of the contemporary investors care most about,if establish a with high efficiency,high precision,high accuracy predictive model of stock price movements,for modern investors there is a very important practical significance.Based on the traditional stock forecasting method and the traditional artificial intelligence forecasting method,this paper trains the improved hidden markov model through the historical stock data,obtains the stock price trend model in recent years,and applies this model to the stock trend prediction in the future.Since the initial state of the traditional hidden markov model is usually generated randomly,it is easy to cause the problems of slow training speed,multiple iterations,local optimization and low accuracy and accuracy.Therefore,in order to solve these problems and improve the training efficiency,reduce the number of iterations and improve the prediction accuracy,this paper proposes an improved genetic algorithm-hidden markov model and prediction strategy.Its main contents include:(1)in the model,the initial state parameters of the hidden markov model are optimized by using genetic algorithm,so as to improve the training efficiency of the hidden markov model,reduce the number of training iterations,and improve the accuracy and accuracy to a certain extent.(2)in terms of strategy,the multi-day weighted average method is used to process stock data,so that the prediction effect of the model is affected by the weight of sudden events,and further improve the accuracy and correctness of the prediction.Finally,the model prediction performance is compared with other traditional prediction models,and the comparison results and analysis are illustrated.Finally,the paper summarizes the whole paper,and proposes the practical significance and application value of the improved model,which provides new research ideas and tools for the control of stock investment risk.