Skip to content

Graph Koopman Autoencoder

Home / Concepts / Graph Koopman Autoencoder

Graph Koopman Autoencoder

Auto-generated stub. Edit this file to add more details.

A GNN-based autoencoder that utilizes Koopman theory to project non-linear spatio-temporal dynamics into a linear subspace for scalable forecasting.

Why It Matters

It provides a novel, reusable mechanism for compressing non-linear, high-dimensional spatio-temporal graph dynamics into a linear subspace for efficient forecasting.

Evidence

a Graph Koopman Autoencoder (GKAE) forecasts spatio-temporal behavior within a linear subspace for each orbital shell.

Metadata & Links