Skip to content

Minimum Description Length Spatial Regionalization

Home / Concepts / Minimum Description Length Spatial Regionalization

Minimum Description Length Spatial Regionalization

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

A framework for partitioning spatial domains into contiguous regions by jointly optimizing for spatial contiguity and compression of temporal archetypes using the Minimum Description Length (MDL) principle.

Why It Matters

This provides a principled, non-parametric, and scalable alternative to ad-hoc spatial regularization for partitioning spatiotemporal data, offering a unique compression-based perspective on clustering.

Evidence

Utilizing the minimum description length principle from information theory, here we propose an efficient and fully nonparametric framework for the regionalization of spatial time series.

Metadata & Links