Paper
8 May 2024 Effects of hyper-parameters in online constrained clustering: a study on animal videos
Francis J. Williams, Ludmila I. Kuncheva
Author Affiliations +
Proceedings Volume 13162, Fourth Symposium on Pattern Recognition and Applications (SPRA 2023); 131620B (2024) https://doi.org/10.1117/12.3030009
Event: Fourth Symposium on Pattern Recognition and Applications (SPRA2023), 2023, Napoli, Italy
Abstract
The aim of online clustering is to discover a structure in running data. Adding label constraints or pairwise constraints to this has shown to improve the clustering accuracy. In this study we present an analysis of how different hyperparameters – proportion of constraints, initial number of clusters, and batch window size – affect most recent and popular online constrained clustering methods, using three different metrics. Our results show that initial number of clusters and window size have an effect on clustering results, while the proportion of constraints does not. We also demonstrate that online clustering performs better than clustering of the whole data together. Our overall findings point at the need for new, more effective online constrained clustering methods.
(2024) Published by SPIE. Downloading of the abstract is permitted for personal use only.
Francis J. Williams and Ludmila I. Kuncheva "Effects of hyper-parameters in online constrained clustering: a study on animal videos", Proc. SPIE 13162, Fourth Symposium on Pattern Recognition and Applications (SPRA 2023), 131620B (8 May 2024); https://doi.org/10.1117/12.3030009
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
Video

Animals

RGB color model

Data processing

MATLAB

Video processing

RELATED CONTENT

YUV chrominance sub-sampling comparison using H.264
Proceedings of SPIE (May 20 2015)
Dali multimedia software library
Proceedings of SPIE (December 14 1998)
Automatic video segmentation and indexing
Proceedings of SPIE (August 26 1999)

Back to Top