Enhancing Stock Trading Performance in Chennai: A Reinforcement Learning and Sentiment Analysis Approach for Real-Time Decision-Making

Trading in stocks is facilitated by both technical and sentiment factors that are not well captured in routine models. The work presented investigates a reinforcement learning model integrated with sentiment analysis for improved stock trading in Chennai. The model analyses live data of stock and sentiment and makes changes to the trades being made dynamically. Training and finally testing the model, a dataset that consisted of stock prices and input sentiment data was used. Different results illustrate an average profit of 12.5% and a maximum drawdown of 5.2% with a Sharpe ratio of 1.8 to outperform previous methods. Profitable trades perfectly solved by the model to a success rate of 78%, and it had great improvements from success rates of 54%, 62 %, and 69 % in the years 2021, 2022, and 2023 respectively. In conclusion, the proposed model's performance suggests that the integration of sentiment analysis and reinforcement learning translates to large improvements in stock trading strategies.