Research And Implementation Of Public Fund Quantification System Based On Deep Learning

With the rapid development of China’s social economy and the listing of the CSI300,CSI 500 and SSE 50 stock indices,the stock trading market is gradually becoming more and more regulated and mature,and the convenient way of stock trading brings huge market trading data.With the rapid development of artificial intelligence technology,the application of deep learning to quantitative investment has become a hot topic of research.Firstly,this paper model the quantitative strategy trading process builds a quantitative stock strategy system and designs a quantitative stock system for public fund investment management based on the multi-factor stock selection strategy trading method.The functional development requirements and overall architecture of the system are introduced in detail,and the programmed automatic investment trading function is implemented to replace the traditional manual trading operation through quantitative trading and avoid the trading process being influenced by subjective experience and market sentiment.To achieve the basic functions of stock data acquisition,data pre-processing,historical data trading and visualization of trading results.To get a better return on investment,based on the theory of strong market effectiveness,the market trend shows an independent and identical distribution pattern within a short period,and the trading parameters are dynamically optimized through rolling training methods to achieve quantitative trading excess returns.The final investment results are compared with the CSI 300 benchmark investment returns,and the corresponding trading indicators are calculated to assess the investment risk and measure the optimization effect of the trading model.Secondly,this paper implements a quantitative trading system based on deep learning,constructs an LSTM deep neural network model to forecast stock prices,improves the lagging problem of the LSTM model in predicting stock data through feature engineering,calculates the mean absolute percentage error(MAPE)to evaluate the prediction ability of the model on stock prices,and at the same time combines the multi-factor stock selection strategy to optimize the trading stock pool and improve the quantitative strategy The system is based on a large volume of public equity stocks.The system is based on the characteristics of large volume and continuous investment of public funds to develop a quantitative intelligent trading system,using B/S structure using mainstream technologies such as PHP and My Sql database,combined with multi-factor stock selection strategy and model prediction function,to achieve the system on the historical data simulation trading,enhance the strategy investment returns,through the system trading verification,to assist users in the stock market investment decisions.