{"repo":"kay-ou/SimTradeML","free":true,"listed":false,"github":"https://github.com/kay-ou/SimTradeML","clone":"git clone https://github.com/kay-ou/SimTradeML.git","description":"SimTradeML is the predictive engine of the SimTrade ecosystem, turning financial data into ready‑to‑use machine learning models. It offers a lightweight training platform that outputs .pkl files fully compatible with Ptrade, seamlessly integrating into strategies in SimTradeLab and Ptrade.","language":"Python","stars":36,"topics":["backtesting","machine-learning"],"license":"AGPL-3.0","category":"machine-learning","readme_excerpt":"SimTradeML English 中文 Deutsch PTrade-compatible quantitative trading ML framework - helps users quickly train prediction models for use in SimTradeLab and PTrade. Core Positioning SimTradeML is the machine learning toolchain for SimTradeLab: - 🎯 Optimized for PTrade : Trained models can be used directly in SimTradeLab backtesting and PTrade live trading - ⚡ Fast Training : From data to usable model in 5 minutes - 📊 Quantitative Finance Metrics : Professional evaluation with IC/ICIR/quantile returns - 🔧 A-share Ecosystem Integration : Deep integration with SimTradeLab data sources Quick Start Installation Train Your First Model in 5 Minutes Complete Examples See examples/ directory: - mvp train.py - Complete training pipeline (data collection, training, export) - complete example.py - Recommended usage demonstration (single-file package) Recommended Usage (Single-File Package) Core Features PTrade Compatibility - ✅ XGBoost 0.90 : PTrade supported version - ✅ Flexible Save Formats : Supports JSON, Pickle, XGBoost native formats - ✅ Plug and Play : Trained models can be used directly in SimTradeLab ML Capabilities - Data Source Abstraction : Easily switch between different data sources - Feature Engineering : Built-in technical indicators, supports custom features - Evaluation Metrics : IC/ICIR/quantile returns/directional accuracy - Parallel Processing : Automatic multi-process sampling acceleration Quantitative Finance Specialization - Time Series Rigor : Prevents future da","default_branch":null,"files":null,"tree":[],"storefront":"/r/kay-ou","claimed":false,"request_supported":{"post":"https://gitbuyer.com/r/kay-ou/SimTradeML/request-supported","requests":0},"note":"indexed from public GitHub; nothing is for sale on this page. Clone it from GitHub. Paid listings live at /search."}