Analysis of Stock Price and Price Movement Prediction based on Machine Learning Models for E-Hualu

Abstract

At present, many multi-information input models are used in various stock forecasts, but for most Chinese investors, it is difficult to obtain effective stock information, and most of them buy and sell stocks only based on stock trends and stock prices. Therefore, in order to get better prediction results based on basic information, the author tries to use some machine learning methods to predict E-Hualu in the Chinese stock market, and the aim is to select the best prediction model. In this study, five machine learning models were applied to predict stock prices, namely linear regression, SVR, random forest, XGBoost, and LSTM, then the close price of the previous 14 days were converted into the input of each model to estimate the close price of the next day. After the price prediction results were obtained, secondary processing was carried out, transforming previous outputs into trend prediction results. The research results indicate that the LSTM model has outstanding performance in both price prediction and trend prediction. Hopefully, the results of this research can provide some suggestions for investing in E-Hualu.