Design And Realization Of Stock Analysis System Based On Improved Bollinger Bands

With the development of the stock market,the stock market has become an essential part of China's economic development,playing a very important role in the country's economy,and more and more people are using stock trading as a way to manage their money.More and more people use stock trading as a way of financial management.With the development of computer technology,the use of computer technology to analyze stock information is more convenient and reliable.Therefore,there are more and more stock analysis systems on the market,but these systems are too complex and specialized,and their operability is poor,leading to many investors cannot obtain all the effective information that affects stock fluctuations in time,and some systems lack intelligent analysis functions to help investors make effective investment decisions.Therefore,it is very important to develop a simple,practical,and fully functional stock analysis system.Aiming at the above problems,this paper realizes a stock analysis system based on Bollinger Bands.The system is designed and implemented based on the B/S structure,using Python and Flask to build the back-end server,and the front-end Web page is developed using Html.This article mainly realizes the basic functions of stock market data browsing,comment management,self-selected stock management,etc.,and focuses on the design and realization of the three functions of similar K-line,improved Bollinger Bands and stock price prediction.First,in order to achieve the similar K-line function,the historical K-line data of all stocks in Shanghai and Shenzhen are divided into sub-sequences,and then the Pearson correlation coefficient is used to calculate the similarity between the recent K-line data of each stock and these sub-sequences,and the calculation results are sorted and selected.Secondly,based on the optimal buying and selling points marked in the historical data,the method of multiple linear regression is used to respectively fit the upper and lower trajectories of the Bollinger Bands to realize the improved Bollinger Bands function.Finally,based on data such as the upper and lower trajectories of the improved Bollinger Bands,the closing price and trading volume of stocks,a long-and short-term memory loop neural network is used to predict stock prices.In addition,according to the different characteristics of the data,this article uses different databases for data storage.Specifically,the MongoDB database is used to store frequently queried data such as stock quotes and news information;while the data with strong relationships such as user information and comment information is stored in the MySQL database;at the same time,in order to improve the operating speed of the system,the Redis middleware is used in the design and implementation of the system in the article.Compared with the traditional Bollinger Bands for trading,the improved Bollinger Bands designed and realized in this paper can achieve a higher rate of return.In addition,compared to only using the closing price to predict the stock price,the multi-dimensional data stock price prediction realized by this system can achieve higher accuracy.And after testing,the functional logic of the system is correct,and it has good performance in terms of interface response speed,throughput and other performance.