Adrian isn't taking new orders for this project right now. Here are some similar projects to explore.
You will get Quantitative Portfolio Optimization with Bayesian ML and Black-Litterman
Project details
Kobayashi - Quantitative Portfolio Optimization System
A production-ready quant trading system combining Bayesian ML with Black-Litterman portfolio optimization.
Live demo link available at:
kobayashi(dot)fit
What Sets This Apart:
• Uncertainty-Aware Predictions: Bayesian models (BayesianRidge, ARD) provide uncertainty estimates. High-uncertainty predictions are auto down-weighted, creating robust portfolios.
• Black-Litterman Integration: ML predictions blend with market equilibrium, preventing extreme bets and gracefully handling model errors.
• Per-Symbol Feature Selection: Each stock gets its own optimized feature set from 200+ technical indicators.
• Clean Architecture: Professional 5-layer design with dependency injection - maintainable, testable, extensible.
• 7 Optimization Strategies: Mean-Variance, HRP, CVaR, CDaR, Semivariance, CLA, Equal Weight.
• Interactive Streamlit UI: Run experiments, track results, compare strategies visually.
• Comprehensive Documentation: 6 architecture guides, 34-question FAQ, ELI5 explanations.
Built with Python 3.11+, scikit-learn, PyPortfolioOpt, VectorBT. Realistic backtesting with transaction costs and slippage.
A production-ready quant trading system combining Bayesian ML with Black-Litterman portfolio optimization.
Live demo link available at:
kobayashi(dot)fit
What Sets This Apart:
• Uncertainty-Aware Predictions: Bayesian models (BayesianRidge, ARD) provide uncertainty estimates. High-uncertainty predictions are auto down-weighted, creating robust portfolios.
• Black-Litterman Integration: ML predictions blend with market equilibrium, preventing extreme bets and gracefully handling model errors.
• Per-Symbol Feature Selection: Each stock gets its own optimized feature set from 200+ technical indicators.
• Clean Architecture: Professional 5-layer design with dependency injection - maintainable, testable, extensible.
• 7 Optimization Strategies: Mean-Variance, HRP, CVaR, CDaR, Semivariance, CLA, Equal Weight.
• Interactive Streamlit UI: Run experiments, track results, compare strategies visually.
• Comprehensive Documentation: 6 architecture guides, 34-question FAQ, ELI5 explanations.
Built with Python 3.11+, scikit-learn, PyPortfolioOpt, VectorBT. Realistic backtesting with transaction costs and slippage.
Machine Learning Tools
NumPy, pandas, Python, Python Scikit-Learn, scikit-learn, SciPyWhat's included $25,000
These options are included with the project scope.
$25,000
- Delivery Time 14 days
- Number of Revisions 0
- Number of Model Variations 2
- Number of Scenarios 1
- Model Validation/Testing
- Model Documentation
- Source Code
Optional add-ons
You can add these on the next page.
Fast 7 Days Delivery
+$1,000
Additional Revision
+$1,000
Additional Scenario
(+ 3 Days)
+$1,000
Additional Graph/Chart
(+ 3 Days)
+$1,000
Data Source Connectivity
(+ 3 Days)
+$1,000About Adrian
AI & Machine Learning
Mosoaia, Romania - 1:35 am local time
My expertise is ranging from Assembly Language(s) & Reverse Engineering to Artificial Intelligence & Algorithmic Trading.
I'll pretty much handle anything you can throw at me.
Steps for completing your project
After purchasing the project, send requirements so Adrian can start the project.
Delivery time starts when Adrian receives requirements from you.
Adrian works on your project following the steps below.
Revisions may occur after the delivery date.
Prepare delivery
The work is (mostly) done, I'll make the last double-checks, and deliver the project within the requested timeframe.
Add-ons
If The Client orders any add-on, I'll start working on it, as soon as possible.