📈 Stock Sentiment Predictor

An advanced machine learning system for stock price prediction that combines technical analysis, sentiment analysis, and state-of-the-art deep learning models.

🚀 Features

Core Capabilities

• 📊 Real-time stock data via yfinance API
• 📰 News sentiment analysis using NewsAPI + VADER
• 🤖 Multiple ML models: Random Forest, XGBoost, LSTM, Transformer
• 📈 Interactive dashboard built with Streamlit
• 🔄 Backtesting framework with trading strategy simulation
• 📄 Automated PDF/Markdown reports

Advanced ML Components

• 🎯 Heterogeneous Ensemble: Combines Random Forest + XGBoost + Ridge with a meta-learner.
• 📉 Uncertainty Quantification: Provides confidence intervals using quantile regression.
• 🔀 Market Regime Detection: Uses HMM-based adaptive models for different market conditions.
• 💬 Sentiment-Adaptive Weighting: Dynamically adjusts feature weights based on news volume.
• 🧠 Deep Learning Models: Includes LSTM, Transformer, and Temporal Fusion Transformer (TFT).

📦 Installation

git clone https://github.com/Akhilesh-0529/stock-sentiment-predictor.git
cd stock-sentiment-predictor

# Create and activate virtual environment
python -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

# Install dependencies
pip install -r requirements.txt

# For Mac users, install OpenMP for XGBoost
brew install libomp

🎯 Quick Start

1. Run Dashboard

streamlit run src/api/simple_dashboard.py

2. Train Models

# Train a model for a single stock
python -c "from src.models.training_pipeline import TrainingPipeline; TrainingPipeline(['AAPL']).train_model('AAPL')"

🏗️ Architecture

The project is structured to separate concerns, from data ingestion to model deployment.

stock-sentiment-predictor/
├── src/
│   ├── data/
│   ├── sentiment/
│   ├── numerical/
│   ├── models/
│   ├── utils/
│   ├── backtest/
│   └── api/
├── models/
└── reports/

⚠️ Disclaimer

This project is for educational and research purposes only. Do not use this system for actual trading without extensive validation and professional advice.

📄 License

This project is licensed under the MIT License.