aiqbench analysis
This research introduces a scalable framework for multi-domain clustering using measure quantization and optimal transport, effectively handling diverse data types like images, audio, and sensors. By minimizing probability metrics such as Sinkhorn divergence, the method achieves superior performance over classical baselines while maintaining efficiency at large scales. The work advances unsupervised learning by providing a robust collaborative strategy that unifies samples across different domains without relying on traditional centroid-based approximations.
Multi-Domain Clustering via Measure Quantization
arXiv:2609.21664v1 Announce Type: new Abstract: Clustering is a fundamental task in data analysis, typically addressed through centroid-based methods such as K-means. In this work, we present a general framework for multi-domain clustering via measure quantization: given samples from multiple domains, we learn a shared set of cluster prototypes by minimizing a probability metric, such as the Sinkhorn divergence or the Maximum Mean Discrepancy, between each domain's probability measure and the measure of prototypes. Data points are then assigned to clusters either via nearest centroid, or via optimal transport, a collaborative strategy that couples all samples within a domain. A mini-batch optimization strategy makes both fitting and assignment scalable, reducing memory and computational cost while preserving clustering performance. Experimental results on 5 multi-domain benchmarks spanning image, audio and sensor data show that our Sinkhorn-based method consistently outperforms classical and multi-domain clustering baselines, and that this advantage persists when scaling to hundreds of thousands of samples.
Abstract:Clustering is a fundamental task in data analysis, typically addressed through centroid-based methods such as K-means. In this work, we present a general framework for multi-domain clustering via measure quantization: given samples from multiple domains, we learn a shared set of cluster prototypes by minimizing a probability metric, such as the Sinkhorn divergence or the Maximum Mean Discrepancy, between each domain's probability measure and the measure of prototypes. Data points are then assigned to clusters either via nearest centroid, or via optimal transport, a collaborative strategy that couples all samples within a domain. A mini-batch optimization strategy makes both fitting and assignment scalable, reducing memory and computational cost while preserving clustering performance. Experimental results on 5 multi-domain benchmarks spanning image, audio and sensor data show that our Sinkhorn-based method consistently outperforms classical and multi-domain clustering baselines, and that this advantage persists when scaling to hundreds of thousands of samples.
Machine Learning (cs.LG); Machine Learning (stat.ML)
arXiv:2609.21664 [cs.LG]
(or arXiv:2609.21664v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2609.21664 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 246 words
Related coverage
Training and Finetuning Multi-Vector Embedding Models with Sentence Transformers
Hugging Face · Aug 26, 2026
[AINews] not much happened today
Latent Space · Oct 3, 2026
[AINews] High Taste Testers Ep. 1
Latent Space · Oct 17, 2026
Multi-Vector (Late Interaction) Embedding Models with Sentence Transformers
Hugging Face · Aug 18, 2026