Comparison of Stock Price Prediction in Context of ARIMA and Random Forest Models

Abstract

With the gradual improvement of the stock market system, coupled with the increase of per capita personal income, speculating in the stock market will become the main way for residents to invest in the future. Being able to predict the return on stocks is a highly desirable thing for investors. Predicting the future trend of stocks is an essential part when investing in stocks. There are many categories of stock prediction. The purpose of this paper is to compare the ability of the ARIMA and the random forest model to predict the stock market by introducing these two models’ theoretical knowledge. The results turn out that the ARIMA model is more suitable for short-term forecasting although it's essentially only valid for linear relationships. As for the random forest model, it has higher accuracy but it's more complex and computationally expensive. Overall, these results shed light on guiding further exploration of stock price prediction based on the state-of-art machine learning scenarios.