Deep Spectral Clustering via Joint Spectral Embedding and Kmeans

Wengang Guo, Wei Ye* · 2024

Read Deep Spectral Clustering via Joint Spectral Embedding and Kmeans by Wengang Guo, Wei Ye* (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-11080.txt.

Description

Wengang Guo and Wei Ye published a research paper in 2024 on a deep learning version of spectral clustering, a method that groups data by mapping it into an embedding space. The classic pipeline builds the embedding and then clusters it in two separate steps, so neither can correct the other. High-dimensional data also strain the similarity graph it needs.

Their design learns the embedding with a neural network and refines the clustering greedily, always fixing the worst-formed cluster direction next. The two parts train together end to end. On seven real data sets the joint method reached top clustering performance, showing the payoff of letting embedding and grouping teach each other.

Catalog record

CREATORWengang Guo, Wei Ye*
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE34 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-11080

Browse the full library · About the Tower of Babel