Predict stock market price by applying ANN, SVM and Random Forest

School of AI and Advanced Computing, Xi’ an Jiaotong Liverpool University, Suzhou, China

^* Corresponding author: Sicheng.Ji23@student.xjtlu.edu.cn

Abstract

In this modern society, stock market has become one of the most significant things for both person and unity, which can make huge influence. Therefore, more and more researchers attempt to invent useful models to assist them to gain benefits. After scholars’ efforts, people discover that machine learning is an effective model to forecast the price, and ANN, SVM, RF are three famous machine learning models. They are always applied for predicting stock products price. This essay will discuss which model can predict price better to provide assistance to investors. The research depends on data in google scholar and uses a website called colab to analysis the effects of these models and find the better model(s). In this paper, gold price and several enterprises’ historical stock price is used to find out what results can these models provide. We will use trained models and the results will be shown by coordinate maps. In the research, we find that RF is the best model when doing prediction, it obvious has better effects. RF can make forecast more accuracy than other two machine learning models. Hence, for those three machine learning models, Random Forest model can provide better assistance when people want to predict stock price. It means that if investors or companies are eager to obtain wealth or have more advantages in stock market, Random Forest model can be a helpful choice. It is recommended to use this model in the region of stock market.