Stock Prediction based on LSTM Neural Network

Abstract

The stock prediction has always been a difficult problem for investors. The neural network has a predictive effect. Therefore, the research content of this paper is stock price prediction based on LSTM neural network model. Data indicators include the opening price, highest price, lowest index, closing price, trading volume and volume of business. After data standardization and dividing the training set and test set, this paper selects the opening price for neural network prediction, adjusts and optimizes the parameters of the model, obtains the loss function indicators that meet expectations, and then incorporates other indicators into the model for prediction, and it can be obtained more accurate stock price data trends.