Deep Learning-Based Forecasting of Adani Power Shares: A Comprehensive Analysis of Time Series Models and Sentiment Analysis to Enhance Predictive Accuracy
摘要
Prediction of stocks and share price plays an important role in the share market but it can be a very challenging task due to the volatility that is faced in stock markets. This research paper focuses on integrating time series models and sentiment analysis to enhance the prediction accuracy. By using two models Light GBM and ARIMAX models, this research work analyzes fourteen years of ADANI stock data, evaluating performance with MSE and MAE. On the other hand, this paper also used web scraping in order to perform sentiment analysis to get better insight of stock selection by people, stock holding duration, and to analyze the stocks. Goal of this research paper is to combine social media sentiments data with trading data and technical indicators that in return improves accuracy in stock predictions.