Clustering
Overview
Clustering seeks useful groupings when labels are absent or incomplete, helping reveal structure in data such as text, images, graphs, time series, and scientific measurements. Methods include centroid-based k-means, density and hierarchical clustering, graph and spectral methods, mixture models, and newer approaches that learn embeddings or use optimal transport; fairness, privacy, streaming, and federated settings add further constraints. Work also studies how to choose the number of clusters and assess stability, uncertainty, interpretability, and statistical significance. No single method works best across data types: deep methods do not consistently beat strong conventional algorithms, and reliable validation remains a central challenge.
Latest papers
370 papers29 papers in the last four weeks, down 9% on the four weeks before. 0.2% of all new papers.
Multi-Domain Clustering via Measure Quantization
Combining Object Detection with Geometry-Aware Clustering to Distinguish Overlapping Plants in UAV Imagery
Randomized SVD Approximations for Spectral Co-Clustering of Word-Document Matrices
Bracketing Uncertainty in Clustering Under the Manifold Hypothesis
A Spatiotemporal Extension of the Neuromorphic DBSCAN Implementation
flat'' and systolic''. The flat'' construction prioritizes throughput, while the systolic'' construction trades time for space resulting in a smaller, more hardware-friendly architecture at the cost of throughput. In this work, we offer spatiotemporal extensions of these two constructions to better leverage the spatiotemporal nature of event sensor data. Moreover, as in our prior work, we discuss partial or segmented implementations that further leverage time for space when hardware resources are constrained. All network constructions are provided as open-source implementations.