AN OPTIMIZED DEEP LEARNING APPROACH TO STOCK PRICE PREDICTION BASED ON INVESTOR SENTIMENT

Abstract

Stock price prediction is a crucial yet challenging task due to the inherent volatility of financial markets. Traditional forecasting methods often struggle to incorporate the influence of investor sentiment, which plays a significant role in price fluctuations. This paper proposes an optimized deep learning approach that integrates historical stock data and investor sentiment analysis to enhance prediction accuracy. The proposed model leverages natural language processing (NLP) techniques to extract sentiment from financial news, social media, and investor discussions, which is then combined with technical indicators. A hybrid deep learning framework incorporating Long Short-Term Memory (LSTM) networks and attention mechanisms is utilized to capture both temporal dependencies and sentiment-driven market movements. The model is further optimized using hyperparameter tuning and feature selection techniques to improve robustness. Experimental results on real-world stock market datasets demonstrate that our approach outperforms traditional machine learning models in terms of prediction accuracy, trend detection, and risk minimization. This research highlights the importance of combining quantitative data with qualitative investor sentiment for more precise and adaptive stock price forecasting.