DeepStockInsight: Hybrid Model for Stock Price Forecasting

Project Overview

DeepStockInsight is a hybrid model designed to predict stock prices by integrating two powerful data sources: historical stock prices and sentiment analysis of news headlines. This project aims to enhance the accuracy of stock price predictions by leveraging traditional numerical analysis combined with sentiment-driven insights from news articles, creating a more robust forecasting model.

---

Features

• Data Integration: Combines historical stock data from BSE SENSEX with sentiment scores derived from Times of India news headlines.
• Sentiment Analysis: Utilizes the VADER sentiment analysis tool to extract sentiment scores (positive, neutral, negative, and compound) from news headlines.
• Predictive Modeling: Employs an LSTM-based RNN model to capture temporal dependencies in stock prices and sentiment data.
• Performance Metrics: Evaluates the model's accuracy using Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and R2 Score.

---

Dataset

Stock Data:

• Source: Yahoo Finance via the yfinance library
• Timeframe: January 2, 2001, to December 31, 2020
• Features: Open, High, Low, Close, Adjusted Close, Volume

News Data:

• Source: Times of India News Headlines dataset
• Timeframe: January 2, 2001, to December 31, 2020
• Features: Publish Date, Headline Text

---

Preprocessing

Stock Data:

• Removed duplicates
• Converted dates to datetime format
• Filtered relevant columns: Date, Close, Open, High, Low, Volume

News Data:

• Removed duplicates
• Converted publish date to datetime format
• Grouped headlines by date

Data Integration:

• Merged stock data with sentiment scores from the news headlines
• Generated sentiment scores for each headline using the VADER tool
• Dropped irrelevant columns and arranged the dataset for modeling

---

Modeling

Architecture:

• LSTM Layers: 3 LSTM layers with 100 units each
• Dropout Layer: Prevented overfitting with a dropout layer
• Dense Layer: Final output layer predicting the stock price

Training:

• 80% of the data used for training, 20% for testing
• Scaled features using MinMaxScaler
• Model trained over 10 epochs with a batch size of 8

Evaluation:

• Training Loss: 0.2717
• Test Loss: 1.2888
• RMSE (Root Mean Square Error): 651.2819
• R2 Score: 0.9738

---

Results

The model successfully predicted stock prices with high accuracy, achieving an R2 score of 0.9738 and an RMSE of 651.2819. The inclusion of sentiment analysis added valuable context to stock price movements, improving prediction performance.

---

Visualization

The predicted stock prices are plotted against the actual test data to visualize the model's performance. The resulting plot is saved as Stock Price Prediction.jpg.

---

Installation

1. Clone the Repository:

   git clone https://github.com/your-username/DeepStockInsight.git
   cd DeepStockInsight

2. Install Dependencies:

   pip install -r requirements.txt

---

Usage

Run the project script or Jupyter notebook to train the model and make predictions. The code will:

1. Load and preprocess stock and news data.
2. Perform sentiment analysis on the news headlines.
3. Train the LSTM model for stock price prediction.
4. Evaluate and visualize the results.

---

Contributing

Contributions to the project are welcome! Please feel free to fork the repository and submit pull requests for any improvements or bug fixes.

1. Fork the repository.
2. Create a new branch.
3. Implement your changes.
4. Submit a pull request with a description of the changes.

---

Happy coding! 😊