Stock Investor's Sentiment Analysis Using Advanced Machine Learning Algorithms

Macroeconomic factors, investor sentiment, and market trends have a role to play in predicting stock prices, as evident in the financial markets. There is complexity, and challenges remain on how to predict investor sentiment better. Here, we address the gap in the conventional predictive models (failing to understand investor sentiment or the psychological aspect of the stock movements) and present the competitive advantage of cutting-edge models, tools, and technology. As a result, we utilized historical stock market data from Yahoo Finance from 2010 to 2018 on financial news and sentiment and incorporated sentiment analysis into stock price prediction models. We evaluated sentiment impact on stock price fluctuations by employing five Machine Learning classifiers, including Support Vector Machine (SVM), Logistic Regression, K-Nearest Neighbors (KNN), Decision Tree algorithm (CART), Random Forest, and Deep Learning Model Long Short-Term Memory (LSTM) networks, and VEDAR for sentiment classification. Key findings (VEDAR sentiment analysis) indicated moderate correlations exist between investor sentiment and stock price movement. Also, LSTM achieved the highest accuracy of 99%, demonstrating its ability to capture financial text effectively. Among all the Machine Learning classifiers, Logistic Regression and Support Vector Machine performed better, achieving 85% and 84% accuracy, respectively. We presented a holistic approach by incorporating sentiment analysis to analyze stock movements from major companies. This would also directly benefit insti-tutional investors, financial analysts, and fintech companies by enhancing their stock price prediction strategies. Finally, traders and retail investors feel more secure and less anxious in their decision-making by predicting better stock prices in the evolving market using sentiment-driven hybrid techniques.