Semi-supervised Spectral Clustering

Xiaoyi Mai 1 Romain Couillet 1, 2
2 GIPSA-CICS - CICS
GIPSA-DIS - Département Images et Signal
Abstract : In this article, we propose a semi-supervised version of spectral clustering, a widespread graph-based unsupervised learning method. The semi-supervised spectral clustering has the advantage of producing consistent classification of data with sufficiently large number of labelled or unlabelled data, unlike classical graph-based semi-supervised methods which are only consistent on labelled data. Theoretical arguments are provided to support the proposition of this novel approach, as well as empirical evidence to confirm the theoretical claims and demonstrate its superiority over other graph-based semi-supervised methods.
Document type :
Conference papers
Complete list of metadatas

https://hal.archives-ouvertes.fr/hal-01982268
Contributor : Romain Couillet <>
Submitted on : Tuesday, January 15, 2019 - 3:18:04 PM
Last modification on : Wednesday, March 20, 2019 - 10:05:26 AM
Long-term archiving on : Tuesday, April 16, 2019 - 3:09:23 PM

File

SSL_clustering(1).pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01982268, version 1

Citation

Xiaoyi Mai, Romain Couillet. Semi-supervised Spectral Clustering. Asilomar Conference on Signals, Systems, and Computers (ACSSC 2018), Oct 2018, Pacific Grove, United States. ⟨hal-01982268⟩

Share

Metrics

Record views

46

Files downloads

69