PySINDy – A package for the sparse identification of nonlinear dynamical systems(github.com)
github.com
PySINDy – A package for the sparse identification of nonlinear dynamical systems
https://github.com/dynamicslab/pysindy
2 comments
The real way to build 'world models'.
Is it applicable to stock trading?
Maybe if you apply it at a microstructural scale (think seconds) with order book and trade print data from many sources/symbols. Certainly not if you blindly apply it to the time series of the price.
I don't see why not - there are papers which use system identification for modelling stocks.