Paper Key : IRJ************933
Author: Praphullakumar Vilarsao Lokhande,Prof. P. R Gadekar
Date Published: 16 Nov 2024
Abstract
This paper presents a novel approach to stock price forecasting by integrating financial news sentiment analysis with ensemble deep learning techniques within a web application. The application leverages React.js for the user interface, Node.js for backend processing, and MongoDB for data management. By combining sentiment data from financial news with historical stock data, this tool aims to enhance prediction accuracy. We demonstrate the effectiveness of ensemble models, combining LSTM, GRU, and Transformer models, showing significant improvement in prediction stability and accuracy. Results indicate that sentiment-enhanced, ensemble-based predictions outperform traditional approaches, providing a robust tool for financial decision-making.
DOI Requested