AI and Data Science in Financial Markets Predictive Modeling for Stock Price Forecasting

Abstract

This study investigates the application of Artificial Intelligence (AI) and Data Science in stock price forecasting, focusing on advanced predictive models such as Long Short-Term Memory (LSTM) networks, Reinforcement Learning (RL), and sentiment analysis. By integrating traditional financial indicators with real-time sentiment data, this research aims to improve the accuracy and adaptability of stock price predictions in dynamic market conditions. Data preprocessing and feature engineering methods were employed to enhance model inputs, while various machine learning and deep learning models were evaluated based on key performance metrics, including Mean Absolute Error (MAE) and Root Mean Square Error (RMSE). The results demonstrated that LSTM and RL models outperformed traditional models, particularly in capturing sequential dependencies and adapting to real-time market changes. The inclusion of sentiment scores provided additional predictive power, underscoring the potential of alternative data sources in financial forecasting. While advanced models showed higher accuracy, they also required substantial computational resources and careful tuning to prevent overfitting. These findings suggest that AI-driven predictive models, when properly integrated and rigorously tested, offer significant advantages in stock price forecasting, particularly for institutional and algorithmic trading applications