👁️ Laplace-Demon (拉普拉斯妖)

"We may regard the present state of the universe as the effect of its past and the cause of its future." — Pierre-Simon Laplace

Laplace-Demon is a physics-informed quantitative research framework designed to predict the technological trends of the NASDAQ-100 index. Built on top of Microsoft's Qlib, it treats tech giants not just as tickers, but as interconnected nodes in a complex, dynamic system.

🚀 Vision: Beyond Traditional Alpha

Traditional quant models focus on low-frequency price-volume data, which is highly efficient and exhausted in US tech stocks. Laplace-Demon shifts the paradigm by introducing:

• Physics-Informed Factors (物理系因子): Utilizing Information Entropy and Kinematic Acceleration to measure the "momentum" of tech capital.
• Geek Alternative Data (极客另类数据): (WIP) Tracking GitHub commit velocity, arXiv AI paper dominance, and Capex-to-compute ratios.
• Laplace Resonance Network (拉普拉斯共振网络): (WIP) A Graph Neural Network mapping the gravitational pull between Silicon Valley tech giants.

🏗️ Architecture

• Data Engine: Downloads and compiles NASDAQ-100 OHLCV data into Qlib's .bin format.
• Demon Core: Custom alpha generation treating market anomalies as localized Laplacian extremums ($\nabla^2 f$).
• Qlib Backend: Leverages Qlib's robust backtesting and model training infrastructure.

🛠️ Quick Start

(Coming soon: Instructions to initialize the demon and run the first baseline backtest on NASDAQ-100.)

---

Created with the vision of reading the Silicon Valley tech tree.