/Data-Mining-on-BTC-Trading-Statistics

Develop about 200 alpha factors from securities report etc, Grid Search/Random Search/Particle Swarm Optimization to improve factors performance

Primary LanguageJupyter NotebookMIT LicenseMIT

This project is mainly about to extract useful features from the time series of BTC trading data, in order to make rational price prediction in the future.

A large portion of effort here is to understand the behavior/property of the financial market by tracking the fluctuation of paraemter values, so I suggest to read the Technical Analysis file, in order to get some basic ideas on what tools people using to analyze the financial data.

First, I have built the function library(.py file) and the method for calculate 191 Alpha factors derived from the Guotai Security report(.ipynb file).