Research map / Learning theory and algorithms / Clustering
Clustering with Hidden Markov Model on Variable Blocks
Lin Lin, Jia Li
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 changed | Where | Measured | Effect |
|---|---|---|---|
| regularize a GMM with a variable block structure (HMM-VB) instead of a single block | modal GMM · Simulation 5.1.1 | CPU time per fit (s): 48.169s -> 4.087s | Improved |
| use HMM-VB with 11 variable blocks instead of a full GMM with 500 components | mouse lung CyTOF clustering · Mouse lung CyTOF | CPU time per model fit (min): 284.2 -> 11.4 | Improved |
| use HMM-VB instead of modal GMM at 5,000,000 samples | large simulated data (d=40) · Simulation 5.1.3 (n=5,000,000) | mean ARI: 0.9513 -> 0.9999 | Improved |
| use HMM-VB instead of modal GMM for detecting rare clusters | large simulated data (d=40) · Simulation 5.1.3 | mean number of clusters found: 4.1 -> 4.9 | Improved |
| use HMM-VB instead of K-means, HC, Mclust and modal GMM | simulation with 16 true clusters · Simulation 5.1.1 | matching of true clusters (rare clusters) | Improved |
| reverse the ordering of variables (M2 versus M1) (ablation) | HMM-VB with 5 single-variable blocks · Simulation 5.1.2 (no block structure) | clustering result | No clear effect |
| use HMM-VB with five (misspecified) blocks instead of a single-block GMM (ablation) | clustering · Simulation 5.1.2 (no block structure) | ARI | Hurt |
| add dimension-reduction-free block modelling (HMM-VB) instead of pcaReduce | single-cell genomics data · Single-cell genomics (300 cells) | cell types correctly identified: 4 -> 6 | Improved |
Closest papers by meaning
- Consistent Model-based Clustering using the Quasi-Bernoulli Stick-breaking Process (JMLR 2023 · similarity 0.85)
- Super Deep Contrastive Information Bottleneck for Multi-modal Clustering (ICML 2025 · similarity 0.83)
- Multi-aspect Self-guided Deep Information Bottleneck for Multi-modal Clustering (AAAI 2025 · similarity 0.82)
- Revisiting Instance-Optimal Cluster Recovery in the Labeled Stochastic Block Model (ICML 2025 · similarity 0.82)
- A Unified Framework for Variable Selection in Model-Based Clustering with Missing Not at Random (NeurIPS 2025 · similarity 0.82)
- Global-Local Dirichlet Processes for Clustering Grouped Data in the Presence of Group-Specific Idiosyncratic Variables (ICML 2025 · similarity 0.82)
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.
