Analysis of Regression Techniques for Stock Market Prediction: A Performance Review

Stock market prediction is a highly challenging and popular field in the world of finance. This manuscript uses the financial data of three big private Indian banks which looks highly promising to current investors. The main objective of this work is to analyse the older share price values of AXIS Bank, HDFC Bank and ICICI Bank which are enlisted in National Stock Exchange India and to predict their future behaviour. Here, authors explore various models like Support Vector Regression, Moving Averages, eXtreme Gradient Boosting Regression, Long Short Term Memory and Linear Regression to forecast future stock prices. Based on the comparative analysis between the different models, it is observed that Linear Regression is highly accurate in predicting future market movement with the least prediction error and provides a significant jump in performance over moving averages.