Exploring the Efficacy of Supervised Learning in Indian Stock Market Prediction: A Comprehensive Analysis

This paper investigates supervised learning strategies with a peculiar attention on the employment of machine learning algorithms in the stock market. Daily, weekly, and monthly prices of the NSE NIFTY 50 and BSE SENSEX indexes from January 1, 2012, to January 1, 2017, and from January 1, 2018, to January 1, 2023, are included in the datasets utilized in this work. The research examined six supervised learning models, including regression in linear form, multilinear regression, k-nearest neighbour, regression with decision trees, regression using random forests, and support vector regression. In certain circumstances, each model offers a significant level of accuracy for the daily, weekly, and monthly time frames. As evidenced by the COVID-19 pandemic, it is essential to grasp that these models’ accuracy is susceptible to sudden changes in geopolitical situations or economic policy. This investigation examines and compares the effectiveness of the aforementioned algorithms. The linear regression, followed by multiple regression as well as support vector regression models outperformed k-nearest neighbour, decision tree regression and random forest regression models respectively under certain circumstances.