Scalable Uncertainty Quantification for Black-Box Density-Based Clustering

We introduce a novel framework for uncertainty quantification in clustering. By combining the martingale posterior paradigm with density-based clustering, uncertainty in the estimated density is naturally propagated to the clustering structure. The approach scales effectively to high-dimensional and irregularly shaped data by leveraging modern neural density estimators and GPU-friendly parallel computation. We establish frequentist consistency guarantees and validate the methodology on synthetic and real data.

Source: arXiv:2603.03188v1 - http://arxiv.org/abs/2603.03188v1 PDF: https://arxiv.org/pdf/2603.03188v1 Original Link: http://arxiv.org/abs/2603.03188v1