Derived new probabilistic heuristics for unsupervised clustering, leveraging methods from signal processing & information theory. Our research removed the need for humans-in-the-loop, enabling automated high-dimensional clustering at varying levels of granularity.