Research On Wave Prediction Based On Hidden Markov Model

Economic development has made people’s demand for investment surge,and traditional investment channels can no longer satisfy the ever-changing investment ideas and investment methods.Stock trading attracts more and more investors because of its high risk and high returns.As the key to stock trading,stock price forecasting has also received continuous attention and research from many scholars.The wave theory has become one of the important theoretical bases for stock investment because of its accuracy in describing the operating laws of the stock market.When investors use the wave theory to assist trading decisions,there is often a problem of ex-post rather than ex-ante or in-process effectiveness.And in these researches on wave theory,there are traditional time series models and machine learning algorithms,but these models cannot well fit the operation of the stock market,and their forecasting accuracy needs to be improved.There is a hidden trend in the operation of the stock market,and the market waves in the trend can be directly observed.The hidden Markov model can represent this process from hidden state to visible state.In this paper,a HMM-based wave prediction model is constructed by combining the Hidden Markov Model with the wave to explore the running law of the wave.The core work of this paper is mainly divided into the following two parts:(1)Research on the construction of wave probability dictionary and similarity measurement: Through the classification of waves,a wave probability dictionary is constructed,and the relevant statistical data are analyzed.By classifying the wave index,it is divided into 96 categories,and the probability of each type of waves can be obtained by counting the frequency of each category of waves.For the case of missing category data,it can be calculated with the help of wave similarity.At the same time,the evaluation of wave grouping effect based on wave similarity proves the rationality of wave classification in this paper.(2)Wave forecasting research: On the basis of the wave probability dictionary,the wave forecasting research is carried out.Based on the data of the wave probability dictionary,the model parameters are calculated,and a wave prediction model based on HMM is constructed.Secondly,by improving the traditional price forecasting method,the wave forecasting algorithm of this paper is proposed.Finally,the accuracy of the wave prediction effect under the optimal model parameters is verified by experiments.