AI/ML experiments for stock screening, anomaly detection, and quant signal generation — inspired by O'Neil, Weinstein, Minervini techniques, and deep learning.
This project is an evolving research and engineering sandbox where I explore algorithmic trading ideas, machine learning models, and signal processing techniques for modern stock market analysis.
The goal of this repository is to prototype, test, and share innovative AI-powered approaches for:
- Stock screening using fundamentals and technicals
- Momentum and trend stage detection (e.g., Stan Weinstein Stage Analysis)
- Quantitative filtering aligned with the CAN SLIM method
- Anomaly detection for early entry points
- Deep learning ranking models and signal generators
It blends classical strategies with cutting-edge ML.
This repo is actively evolving. I'm currently organizing and sharing code from past experiments (including my Quant Project used during the U.S. Investing Championship), and building new modules with:
- 📊 Custom stock sorter combining O'Neil + Weinstein + Minervini rules
- 🧠 Feature engineering and signal scoring with ML
- 📈 Exploratory notebooks and visual dashboards
🚧 Expect ongoing commits and improvements. I’ll be sharing what I have progressively.
This project is a personal R&D initiative rooted in my competitive trading strategies and long-term vision to build intelligent financial systems. It complements my core professional focus on AI/ML for RF and microwave engineering.
Eventually, the most effective tools and models here may evolve into:
- Real-time signal APIs
- Dashboard-based trading assistants
- Proprietary trading system backends
quant-lab/
├── projects/
│ ├── stock_sorter_v1/ # First full ML stock screening system
│ └── ... (future models)
├── data/ # Sample inputs, watchlists, filtered stocks
├── requirements.txt
├── README.md
└── LICENSE