Enhancing Intraday Trading through Machine Learning: A NSE Nifty 50 Analysis

This paper explores the intersection of machine learning (ML), intraday trading, and the economic landscape of India, focusing on the National Stock Exchange (NSE) Nifty 50 index. The study evaluates various ML models, including K-Nearest Neighbors (KNN), Random Forest Regressor (RF), Decision Tree Regressor, Support Vector Regression (SVR), and Linear Regression (LR), revealing LR as the most proficient model with a remarkable R2 score exceeding 0.997 across CLOSE, HIGH, and LOW values. A key observation is that traditional ML techniques primarily rely on historical data for predictions. However, for intraday strategies, the incorporation of OPEN prices on the day of prediction proves to be a valuable addition. Upon closer examination, the approach utilizing combined data, utilizing both OPEN prices and historical data, demonstrates superior performance with the highest R2 scores (0.99940 and 0.99903) and lowest errors in the HIGH and LOW cases. Conversely, the second approach, excluding OPEN prices, exhibits the least error and the highest R2 score (0.99998) for CLOSE price prediction. In conclusion, incorporating OPEN data in intraday strategies emerges as a compelling and insightful method, significantly enhancing predictive accuracy. The findings highlight the pivotal role of ML in predicting stock prices within India’s NSE Nifty 50, not only providing accurate forecasts but also optimizing intraday trading decisions. As financial markets continue to evolve, the integration of ML and intraday strategies emerges as a promising avenue for achieving more informed and data-driven investment practices in the economic landscape.