Skip to content

Utilizing the sophisticated Recurrent Neural Network (RNN) models, particularly Long Short-Term Memory (LSTM), in combination with FinBERT sentiment analysis, to improve the precision of stock portfolio analysis.

Notifications You must be signed in to change notification settings

mahim37/Dynamic_Portfolio_Analyst

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

42 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Dynamic_Portfolio_Analyst

The stock market's role in the economy is crucial, often signaling prosperity or recession. Accurately predicting short-term movements in the market is a challenging yet rewarding endeavor for investors. This project showcases a research study aimed at advancing stock portfolio using deep learning networks and sentiment analysis derived from financial, business, and technology news articles.

Objective

The primary goal of this research is to leverage complex Recurrent Neural Network (RNN) models, specifically Long Short-Term Memory (LSTM), and FinBERT sentiment analysis to analyze stock prices more accurately. By integrating sentiment analysis from news articles into the LSTM model, we seek to enhance the accuracy of short-term market predictions.

Methodology

The study involves utilizing web scraping techniques to collect financial news articles and historical stock price data. The analysis primarily revolves around employing LSTM models in conjunction with FinBERT sentiment analysis, creating a synergy between deep learning, sentiment analysis, and the financial market.

Key Findings

Our findings indicate a significant potential for sentiment-driven predictions to improve stock forecasting accuracy. This research sheds light on the influence of qualitative factors, such as market sentiment, on stock movements, while also recognizing the challenges and opportunities inherent in using diverse data sources for market prediction.

Implementation

The project includes code for implementing the LSTM models, FinBERT sentiment analysis, and a user-friendly web interface presenting combined insights for investors, traders, and financial analysts.

Usage

To get started with this dynamic portfolio analyst, follow these steps:

  1. Clone this repository
git clone https://github.com/yourusername/dynamic-portfolio-analyst.git
  1. Running backend server:
cd Backend/
  1. Install the required python packages
pip install -r requirements.txt
  1. Run the backend server by default on localhost:5000
python3 app.py
  1. Running frontend server: Open new terminal in the project root directory
  cd Frontend/
  1. Install the dependencies
 npm install
  1. Finally, run frontend by default on localhost:3000
npm start

Conclusion

This project contributes to the ongoing pursuit of predicting stock market movements through the integration of deep learning and sentiment analysis. It aims to provide valuable insights and tools for market participants while acknowledging the complexities and potential in utilizing diverse data sources for enhanced market prediction.

Future Work

  • Exploring further enhancements in sentiment-driven predictions
  • Addressing challenges and expanding the scope of data sources for improved accuracy

Feel free to explore the code, findings, and implementation details within the project.

About

Utilizing the sophisticated Recurrent Neural Network (RNN) models, particularly Long Short-Term Memory (LSTM), in combination with FinBERT sentiment analysis, to improve the precision of stock portfolio analysis.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published