A Geometry-Aware Framework for Clustering Cylindrical Data
Abstract
Cylindrical data pair an angle with a linear measurement. Clustering that ignores the periodicity of the angle breaks up groups lying across its origin. We formulate the K-means algorithm for a generic distance on the cylinder and instantiate it with the chordal distance of the ambient space and the geodesic distance along the surface, so that the two versions differ in the metric alone. Centroids are exact Frechet means computed with known tools: the mean direction for the chord, the circular Frechet mean for the arc. Seeded by K-means++, the algorithm converges in finitely many iterations at a cost comparable to classical K-means. Over an extensive empirical analysis on simulated data, Euclidean K-means is never meaningfully better and breaks down when clusters cross the origin; the two versions agree for concentrated angles, the chord incidentally prevailing only when the angle is uninformative; a model-based cylindrical mixture is the most variable method but describes elongated, correlated clusters better. In hue-value color image quantization, the cylindrical versions performs better on object detection based on color segmentation; on wind direction and nitrogen oxides data they recover two pollution regimes that Euclidean K-means cuts apart.
Related
- Geometry of Relaxed Fair Regression: A Unified Framework for Aware and Unaware Settings
- Scalable Distortion-Aware Clustering for Fronthaul-Limited Cell-Free MIMO Networks
- Reliable Sample-Level Quantum Error Mitigation via Dominance-Aware Clustering
- Geometry-Adaptive Clustering on High-Dimensional Tori
- Consistent Spatial Clustering of Complex Objects via Bregman Geometry
BibTeX
@misc{pandolfo2026geometryaware,
title = {{A Geometry-Aware Framework for Clustering Cylindrical Data}},
author = {Giuseppe Pandolfo and Luca Coraggio and Antonio D'Ambrosio},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.26321},
eprint = {2609.26321},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.26321},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - UNPB
TI - A Geometry-Aware Framework for Clustering Cylindrical Data
AU - Giuseppe Pandolfo
AU - Luca Coraggio
AU - Antonio D'Ambrosio
PY - 2026
DA - 2026/09/22
PB - arXiv
DO - 10.48550/arxiv.2609.26321
UR - https://arxiv.org/abs/2609.26321
AB - Cylindrical data pair an angle with a linear measurement. Clustering that ignores the periodicity of the angle breaks up groups lying across its origin. We formulate the K-means algorithm for a generic distance on the cylinder and instantiate it with the chordal distance of the ambient space and the geodesic distance along the surface, so that the two versions differ in the metric alone. Centroids are exact Frechet means computed with known tools: the mean direction for the chord, the circular Frechet mean for the arc. Seeded by K-means++, the algorithm converges in finitely many iterations at a cost comparable to classical K-means. Over an extensive empirical analysis on simulated data, Euclidean K-means is never meaningfully better and breaks down when clusters cross the origin; the two versions agree for concentrated angles, the chord incidentally prevailing only when the angle is uninformative; a model-based cylindrical mixture is the most variable method but describes elongated, correlated clusters better. In hue-value color image quantization, the cylindrical versions performs better on object detection based on color segmentation; on wind direction and nitrogen oxides data they recover two pollution regimes that Euclidean K-means cuts apart.
ER -