Finance · ML
Trade Simulator
Solo project

Real-time trade analytics: an online-learning slippage model, Almgren-Chriss market impact and a maker/taker classifier.
Key figures
By the numbers
- 3models in one live pipeline
The problem
Estimate the cost of a trade before placing it, in real time.
The approach
An online-learning SGD model predicts slippage tick by tick, an Almgren-Chriss model estimates market impact, and a logistic-regression classifier labels a trade as maker or taker. All three run in one live pipeline.
The result
A real-time analysis app documented model by model, with the parameters and the reasoning behind each choice.
Built with
- Python
- scikit-learn