Bachelor's thesis on reward function design in RL-based BTC grid trading (PPO, Prospect Theory, CPCV)
-
Updated
May 25, 2026 - HTML
Bachelor's thesis on reward function design in RL-based BTC grid trading (PPO, Prospect Theory, CPCV)
RR-Agent — 自研 A股量化研究工作台 · 因子库 · ML 选股 · CPCV/DSR 回测 · 组合优化 · 算法执行。Self-developed quantitative research workbench for China A-shares: in-house factor library, ML stock selection, CPCV+DSR-validated backtesting, industry-neutral portfolio optimization. Built on multi-source-validated ReachRich data. 不构成投资建议。
Living technical reference for Disuza Quantitative — private quantitative research laboratory, Madrid, Spain. Architecture, anti-overfit methodology (CPCV / DSR / PBO per López de Prado), regulatory posture. Source code proprietary.
Intra-bar-fidelity crypto backtester with overfit gates built in
Profile README
Honest backtest harness with a measured detector: leakage-safe CPCV, Deflated Sharpe, PBO, realistic costs, mechanical PASS/KILL verdict. Strategy audit CLI + verdict API.
Local-first backtest framework for Binance USDT-M perpetual futures with cost-aware simulation, gate-filtered validation, and a five-tier verdict.
Agentic multi-strategy hedge fund: PatchTST forecasts, 4-agent LangGraph debate, CPCV-OOS + DSR validation, HRP with Ledoit-Wolf shrinkage. 10-year OOS Sharpe=0.766.
To associate your repository with the cpcv topic, visit your repo's landing page and select "manage topics."