Forums
•3d ago
pCHA - Forecasting chaotic systems from limited data points
I built pCHA, a closed-source framework for forecasting chaotic dynamical systems from limited data (n=15), with no equations provided.
It maps trajectories into a p-adic ultrametric space (Q_p ) using categorical morphisms, which keeps the error growth linear rather than exponential. It passes 116/117 on dysts benchmark.
Since the code is proprietary, all verification goes through a strict public Blind Test Protocol.
Every result, including failures, is published.
Repository and benchmarks: