“Stock Price Prediction Using Machine Learning and Data Mining Techniques: A Comparative Study on Market Forecasting Models”

Abstract

Predicting stock market trends is a challenging task due to its dynamic, nonlinear, and volatile nature. In this study, various machine learning and data mining algorithms are employed to forecast stock prices based on historical and technical indicators. The research explores the performance of supervised models such as Support Vector Machines (SVM), Random Forests (RF), and Gradient Boosting, along with deep learning models like Long Short-Term Memory (LSTM) networks. Data preprocessing techniques including normalization, feature extraction, and correlation-based feature selection are applied to improve model accuracy. The proposed framework also incorporates sentiment analysis from financial news to enhance prediction reliability. Experimental results conducted on benchmark datasets (NSE and NASDAQ) demonstrate that hybrid LSTM models outperform traditional approaches with an accuracy improvement of 10–15%. The study concludes that integrating data mining with advanced machine learning provides a robust solution for real-time financial forecasting.