AI Paper Atlas

Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes

Bayesian Sparse Gaussian Mixture Model for Clustering in High Dimensions

Dapeng Yao, Fangzheng Xie, Yanxun Xu

JMLR 2025 · Probabilistic and scientific modeling › Uncertainty and Gaussian processes · approach: Structured Gaussian process models

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
use Bayesian sparse GMM instead of PCA-KM, MClust, SKM, CHIMEBayesian sparse GMM · Scenario I (K*=3, s=6)ARI: 0.84 vs 0.54 (PCA-KM)Improved
increase support size s from 6 to 12 (ablation)Bayesian sparse GMM · Scenario I (K*=3)ARI: 0.84 -> 0.98Improved
increase support size s from 6 to 12 (ablation)Bayesian sparse GMM · Scenario I (K*=5)ARI: 0.94 -> 0.99Improved
adaptive estimation of K (Bayesian) instead of silhouette/BIC selectionScenario I K*=3 · Scenario I (K*=3, s=12)replicates with correct K: 98 of 100 replicates vs all competitors estimating 2 clustersImproved
use Bayesian method in presence of a tiny clusterBayesian sparse GMM · Scenario II (small cluster)correct K: K = 3 in 96 out of 100 replicates; competitors K = 2 in all 100Improved
fix K to truth for competitors in Scenario IISKM, PCA-KM, MClust · Scenario II (small cluster, K fixed to 3)ARI: 0.76, 0.84, 0.79 vs 0.99 for BayesianHurt
Gaussian working likelihood under multivariate t data (misspecification)Bayesian sparse GMM · Scenario IIIARI: average ARI of 0.97; K=3 in 76 of 100 replicatesNo clear effect
Gaussian MClust/CHIME under t misspecificationMClust, CHIME · Scenario IIIARI: only one cluster in 93 out of 100 replicates, ARI less than 0.05Hurt
Bayesian method vs K-means variants on real scRNA-SeqBayesian sparse GMM · scRNA-SeqARI: 0.84 vs 0.79 (KM)Improved
Bayesian method vs SKM on scRNA-SeqBayesian sparse GMM · scRNA-SeqNMI: 0.80 vs 0.22 (SKM)Improved

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