Research On Stock Price Trend Prediction Method Based On Financial Data Informatio

As computer technology continues to advance,investors and financial institutions are increasingly interested in accurately predicting stock price trends based on financial data information to reduce investment risks and achieve objective gains.However,the stock market is a dynamic,complex,and chaotic environment,and price fluctuations are influenced by various factors,making it difficult to analyze financial data information about stocks.Therefore,designing a stock price trend prediction model by effectively mining the potential content in financial data information is of great theoretical significance and research value.The current stock price trend prediction models suffer from the issue of being unable to extract useful information from financial data across different stock markets.This problem affects the accuracy of the model in predicting stock prices as it cannot effectively analyze financial data information during the prediction process.In response to this issue,this paper proposes two different types of stock price trend prediction models based on deep learning and reinforcement learning,respectively suitable for the index stock market and individual stock market.The main research content of this paper includes:(1)This paper deeply explores the definition and classification of financial data information,and conducts a systematic study on the impact of stock price trend prediction.At the same time,this paper analyzes the existing stock price trend forecasting methods,and summarizes the scope and limitations of these methods.Finally,this paper introduces the evaluation indicators of commonly used stock price trend forecasting methods to ensure the objectivity and scientificity of the evaluation results.(2)In the index stock market,the existing forecasting methods fail to effectively use the internal relationship between various financial data information,resulting in the model being unable to accurately capture stock price trends.To address this problem,this paper proposes a prediction method based on optimized tensors and deep long short-term memory networks.This method uses tensor modeling of financial data information and designs an optimization algorithm named TOS to preserve the internal relationship between various financial data information.At the same time,the method also uses the TLSTM-LS module designed based on the optimization tensor and the SLSQP optimization algorithm to further strengthen this internal connection.Finally,the method stacks TLSTM-LS modules through designed-based link gates to enable information transfer between different modules to predict exponential price trends.(3)In the individual stock market,the existing forecasting methods cannot effectively analyze and provide timely feedback on various financial data information and their changes in the market,resulting in deviations in the forecast results of the model.To address this problem,this paper proposes a prediction method based on growing neural gas(GNG)and Triple Q-Learning.The method,to effectively parse the financial data information existing in the individual stock market,firstly captures the stock price trend from historical stock data and constructs the trading environment state of the market by the GNG algorithm in unsupervised learning.Secondly,the redesigned reward function RF-S enables the model to provide timely feedback on the trading information existing in the individual stock trading market.Finally,a new trading agent algorithm,Triple Q-learning,is designed based on reinforcement learning to execute corresponding trading behaviors based on the environment state constructed by GNG and to make comprehensive predictions of stock price trends in the individual stock market.