AI Paper Atlas

Research map / Graph learning / Graph self-supervised learning

Deep Clustering by Gaussian Mixture Variational Autoencoders With Graph Embedding

Linxiao Yang, Ngai-Man Cheung, Jiaying Li, Jun Fang

ICCV 2019 · Graph learning › Graph self-supervised learning · approach: Semi-supervised node representation learning · 161 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add graph embedding (JS divergence between neighbor posteriors) to Gaussian mixture VAEVaDE · MNISTclustering accuracy: 94.46 -> 97.58Improved
add graph embedding to Gaussian mixture VAEVaDE · STL-10clustering accuracy: 84.45 -> 90.59Improved
add graph embedding to Gaussian mixture VAEVaDE · Reutersclustering accuracy: 79.83 -> 82.3Improved
add graph embedding to Gaussian mixture VAEVaDE · HHARclustering accuracy: 84.46 -> 89.04Improved
use jointly learned GMM clustering instead of two-stage spectral embedding then k-meansSpectralNet · MNISTclustering accuracy: 97.1 -> 97.58Improved
use DGG instead of IMSATIMSAT · MNISTclustering accuracy: 98.4 -> 97.58Hurt
use DGG instead of IMSATIMSAT · Reutersclustering accuracy: 71.0 -> 82.3Improved
add 1 nearest neighbor of graph information (ablation)DGG · MNISTclustering accuracy: 94.82 -> 96.98Improved
increase neighbors from 1 to 20 (ablation)DGG · MNISTclustering accuracy: 96.98 -> 97.58Improved
increase neighbors from 20 to 30 (ablation)DGG · MNISTclustering accuracy: 97.58 -> 97.49Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy