Quantitative strategy researcher · Python 数学背景的量化策略研究者
I come from a mathematics background and work on systematic strategy research. I also enjoy building small, self-hosted tools that run on my own machines.
数学背景,做系统化的策略研究;业余喜欢造一些能跑在自己机器上的小工具。
📍 Nanjing, China · 💼 LinkedIn
🔭 Focus / 方向
- 📈 Quantitative strategy research — 量化策略研究
- 🧪 Backtesting · factor & signal exploration — 回测、因子与信号探索
- 🎲 Stochastic simulation & option pricing — 随机模拟与期权定价
- 🛠️ Self-hosted Python tools — 自托管 Python 工具
🛠️ Stack
Python · NumPy · pandas · Django · FastAPI · WebSocket · Docker · uv · Git
📌 Selected projects / 代表项目
- 📈 backtest-lab — 轻量级向量化回测框架:信号 T+1 生效(杜绝未来函数)、按换手计提成本,输出净值曲线与绩效指标(夏普 / 最大回撤 / 卡玛)。
A lightweight vectorized backtesting framework — no look-ahead, turnover-based costs, equity curve & risk metrics. - 🎲 monte-carlo — 蒙特卡洛模拟:精确模拟 GBM、Black–Scholes 对拍、收敛性验证(SE ∝ 1/√N)与对偶变量法方差缩减。每个估计都带标准误。
Monte Carlo simulation — GBM paths, option pricing vs. Black–Scholes, convergence & antithetic variance reduction. - 📓 quant-notes — 量化研究背后的数学与统计笔记,与上面两个仓库互为表里:notes 讲「为什么」,代码给「怎么实现」。
Notes on the math & statistics behind quant research — the "why" behind the code. - 🃏 webpokerdealer — 家庭德州扑克发牌助手:服务端权威牌堆 + 按角色过滤视图,手机看底牌、平板当公牌桌。FastAPI + WebSocket + 单容器 Docker。
Self-hosted Texas Hold'em dealer — server-authoritative deck, per-role state filtering. - 🧰 project-scaffold — 项目脚手架:每个新仓库自带 ROADMAP / DECISIONS / AGENTS 文档,做到「仓库自描述」。
Every new repo starts with a ROADMAP / DECISIONS / AGENTS doc set.