A Research on Stock Index Prediction Based on Multiple Linear Regression and ELM Neural Network

摘要

The closing price of the stock index has the characteristics of non-linearity, volatility and noise, and it is difficult to predict. In this paper, the prediction result of the multiple linear regression model is used as one of the ELM model’s input layer vectors, and a stock index prediction model combines the multiple linear regression model and the ELM neural network model. Then, we use the model to predict and analyse the Shanghai and Shenzhen 300 Index. The prediction results show that the combination model has improved prediction and accuracy compared with a single linear model or ELM neural network model. The combination model is practical.