Back to Publications

Geometric Partition Entropy: Coarse-Graining a Continuous State Space

Entropy, 24(10):1432 2022 Open Access

Authors: C. Tyler Diggans and A. R. AlMomani

Introduces Geometric Partition Entropy (GPE), a method for computing entropy of continuous random variables by leveraging the intrinsic geometry of the data through Delaunay tessellation. GPE eliminates the need for arbitrary binning or bandwidth choices, instead deriving partitions directly from the spatial structure of observations.

Related Expertise

Explainable & Interpretable ML / AI (Pillar 2)