Investor Sentiment On Stock Market Prediction

The stock market is a very important part of the financial market,it can reflect the economic operation of the country,often referred to as the economic "barometer","weathervane".It is of great significance for the government and investors to correctly predict the trend of the stock market.The trend of the stock market will be affected by many factors,modern financial market financial scholars rarely consider behavior factors,psychological factors and other factors will affect the economic theme of personal decision-making problem.But in fact the economic subject will be affected by various factors when making decisions,behavioral finance,psychology and behavior will affect the stock market trend of irrational investors in stock market,the stock price from its own value.Many scholars at home and abroad have studied the stock index prediction,but the stock market analysis based on behavioral finance is relatively less.In the background of big data era,this paper uses crawler technology to collect the financial comment information on the network,and uses these text information and the market index data to analyze the financial comment behavior of investors.At first,this paper uses the text processing technology to deal with the financial comment information,studies the investor's financial comment behavior,and carries on the hypothesis verification,and proves the validity of the data.The investor sentiment is studied in this paper by using financial review sentiment classification technology,the financial data of text classification and the sentiment analysis,to quantify investor sentiment and further use of sentiment analysis as auxiliary features predict the stock market future.This paper uses the time-varying probability density function model as the basic model to predict the trend of the stock market,and the results can better fit the market curve,but there is still room for improvement.Then the exploration of using text analysis online stock build emotional factors as variables into time-varying probability density function model.The empirical conclusions are obtained as follows:1)time-varying probability density function model is a non-parameter model has a good prediction effect;2)joined the emotional factor in time-varying probability density function model in the evaluation of the loss function,the result is better than the time-varying probability density function model to forecast effect satisfactory;3)we can infer that the investor's individual behavior and psychological factors will affect the financial activities.The experimental results show that it has a very important guiding significance to the trend of investor sentiment on the stock market,and to raise the stock price prediction accuracy according to investor sentiment index,and then confirms the behavioral finance theory and thought.