This repository contains the implementation for the manifold regression framework presented in our paper. In particular, it can be used to run the experiments (forecasting synthetical spherical trajectories) from the paper.
Esfandiar Nava-Yazdani:
Ridge Regression on Riemannian Manifolds for Time-Series Prediction
Journal of Information Geometry, 2026.
You can run the synthetic validation tests directly in your browser (ensure you clone the repository and add
the root directory to your system path to enable the modules):
Note: If you encounter import errors, try Runtime → Disconnect and delete runtime, then re-run cells to get a fresh environment.
You can also launch the notebook in a fully reproducible environment via Binder (no setup required):
To run the notebooks, ensure you have the dependencies installed:
pip install -r requirements.txt