The Minimum Code Length for Clustering Using the Gray Code

We propose new approaches to exploit compression algorithms for clustering numerical data. Our first contribution is to design a measure that can score the quality of a given clustering result under t