Analysis And Application Of Hidden Markov Model Based On Volatility In Financial Data

The analysis and research on financial data is an important part of the research in the financial field.It aims to explore the inherent laws behind the changes of financial data through analysis,and establish a reliable model to analyze and predict the financial market,so as to reduce market risks and provide effective transaction information for participants in the financial market.Due to the complexity of the financial market,it is difficult for the existing methods to obtain more accurate prediction results.Based on the volatility of stock price,this paper establishes the Hidden Markov Model(HMM)to predict the rise and fall of stock price and stock price.The main work of this paper is as follows:1)A discrete HMM prediction method based on volatility is proposed.By discretizing the stock volatility sequence,the problem that the probability density function of the observed value is difficult to be determined in the traditional HMM when the input observation sequence is a continuous value is solved.In the prediction stage,different from the previous method,this paper combines state transition matrix and probability distribution matrix of observed values to calculate the probability distribution of predicted values by weighting method,so as to get more accurate prediction results.Compared with the existing prediction methods,the prediction method proposed in this paper not only improves the accuracy of prediction results,but also makes the prediction results more explanatory and reliable.2)A HMM prediction method based on speech recognition is proposed.This method introduces the application of HMM in the speech recognition field to the stock rise and fall prediction.To solve the problem that the traditional prediction method does not have a high accuracy in predicting the future rise and fall of stocks,this method divides the stock volatility sequence into several sub-sequences,then marks the positive and negative samples according to the volatility,and uses clustering algorithm to extract the core positive and negative samples to reduce the number of training samples.Then,a HMM is constructed for each core sample and classified into a rising model or a falling model according to the mark.Finally,the probability of the input observation sequence in the two models is used to predict the rise and fall of the stock.Experimental results show that this method can improve the accuracy of stock rise and fall prediction.The two prediction methods proposed in this paper not only improve the existing prediction methods,but also improve the accuracy of stock price and stock rise and fall prediction.The experimental results show that the method proposed in this paper can provide investors with effective trading information,make correct decisions and obtain more economic benefits.By introducing HMM into the analysis of financial data in the field of speech recognition,new ideas are opened for the following work.In future work,we will focus on the parameter optimization of HMM and the selection of stock feature factors,so as to further improve the accuracy of prediction.