VINCE — ML Strategy Optimizer Skill

What problem does it solve?

VINCE provides a machine learning strategy optimization framework for the Four Pillars backtester, enabling rigorous parameter tuning and deep diagnostic analysis to improve trade outcomes.

Core Features & Use Cases

• Bayesian optimization with Optuna for efficient hyperparameter search
• Feature importance assessment with XGBoost and PyTorch-based models
• GPU-accelerated training and local data handling for 370+ coins
• Walk-forward methodology with out-of-sample validation and MAE/MFE/ETD analysis
• Live visualization and dashboards via Streamlit for decision support

Quick Start

Run VINCE to initialize a walk-forward optimization on your strategy and inspect MFE/MAE results.