Study of Linear Regression Prediction Model for American Stock Market Prediction

Abstract

A difficult task in the financial market is to make accurate stock market predictions. This is because there are plenty of factors that affect the price of a company's stock. Even a single bad tweet can affect the price of a stock. As a result, making an exact prediction is a difficult task. Many scientists are working to find a solution that can withstand a wide range of factors and still provide an accurate result. Supervised learning model called linear regression have produced excellent predictions in a variety of fields over the last few decades. In these studies, a model is designed to predict the stock market prediction by using linear regression. The model is evaluated by using the dataset of three best companies that is (Walmart, Tesla, and Amazon) listed on the American stock exchange called NASDAQ and the result is analyzed in terms of root mean squared error. Then the following results are compared with other machine learning models that is Random Forest and Support Vector Machine. In all cases, linear regression gives the best results.