Research on Stock Return Forecasting Methods based on Time Series Models

Accurately predicting the trend of stock return rate is a hot research issue. With development artificial intelligence, machine learning, big data and other technologies, it brings new potential to prediction market. In order accurately predict return, this paper mainly constructs time series ARMA model random forest model, uses stacking method fuse models, predicts daily Yangtze River Electric Power stock. The final fusion has an MSE 1.757 on training set 1.274 test set. overall error within acceptable range. At same time, fused can weaken problem underfitting single which provides valuable reference for optimization research.