Application of LSTM Model based on Particle Swarm Optimization Algorithm in Stock Market Trend Prediction

The stock index reflects the development of national economy and is a dynamic, complex, nonlinear system. Accurately predicting future trends in the stock market has important practical significance for regulators, investment institutions, and individuals. With the rapid development of machine learning and deep learning, using time series models, neural networks, machine learning algorithms and other models to predict future trends in the stock market has become a hot research topic. Although Long Short-Term Memory (LSTM) network model has good memory characteristics and is suitable for predicting stock time series data, it lacks the ability to explain final decisions obtained by the model as well as being affected by human factors. Therefore, this study combines particle swarm optimization algorithm with LSTM model to take optimal position of particles as optimal parameters for LSTM neural network prediction of stock data in order to improve accuracy.