Research map / Learning theory and algorithms / Clustering
Flexible High-Dimensional Unsupervised Learning with Missing Data
Yuhong Wei, Yang Tang, Paul D. McNicholas
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 |
|---|---|---|---|
| model missing values inside the EM algorithm instead of mean imputation before MGHFA | MGHFA · Simulation pattern 1, missing rate 30% | ARI: 0.69 -> 0.90 | Improved |
| model missing values inside the EM algorithm instead of mean imputation before PGMM | PGMM · Simulation pattern 1, missing rate 30% | ARI: 0.32 -> 0.90 | Improved |
| model missing values inside the EM algorithm instead of mean imputation before PGMM | PGMM · Simulation pattern 2, missing rate 30% | ARI: 0.15 -> 0.89 | Improved |
| use mean imputation at low missingness | MI-PGMM and MI-MGHFA · Simulation, missing rate 5% and 10% | ARI: 0.97 -> 0.99 | No clear effect |
| use generalized hyperbolic instead of skew-t factor analyzers with missing data | MSTFAMISS · Simulation pattern 1, missing rate 30% | ARI: 0.89 -> 0.90 | No clear effect |
| use model-based MGHFAMISS instead of k-POD | k-POD · Simulation pattern 1, missing rate 30% | ARI: 0.62 -> 0.90 | Improved |
| use model-based MGHFAMISS instead of k-POD | k-POD · Simulation pattern 1, missing rate 30% | ERR: 0.14 -> 0.04 | Improved |
| use conditional-predictor imputation instead of mean imputation | MI-PGMM · Imputation pattern 1, missing rate 5% | imputation MSE: 30.14 -> 8.70 | Improved |
| use MGHFAMISS instead of MSTFAMISS | MSTFAMISS · Imputation pattern 1, missing rate 20% | imputation MSE: 9.58 -> 9.78 | Hurt |
| use MGHFAMISS instead of MSTFAMISS | MSTFAMISS · Wine data original, missing rate 5% | ARI: 0.82 -> 0.85 | Improved |
| use MGHFAMISS instead of MSTFAMISS | MSTFAMISS · Wine data original, missing rate 30% | ARI: 0.76 -> 0.75 | Hurt |
| add 17 noise attributes to the wine data | MGHFAMISS · Wine data, missing rate 20% | ARI: 0.77 -> 0.72 | Hurt |
Closest papers by meaning
- Random features models: a way to study the success of naive imputation (ICML 2024 · similarity 0.83)
- A Unified Framework for Variable Selection in Model-Based Clustering with Missing Not at Random (NeurIPS 2025 · similarity 0.83)
- Heterogeneous Sufficient Dimension Reduction and Subspace Clustering (ICML 2025 · similarity 0.83)
- Rethinking the Diffusion Models for Missing Data Imputation: A Gradient Flow Perspective (NeurIPS 2024 · similarity 0.83)
- Generative Modeling under Non-Monotone MAR Missingness via Approximate Wasserstein Gradient Flows (NeurIPS 2026 · similarity 0.82)
- The Breakdown of Gaussian Universality in Classification of High-dimensional Linear Factor Mixtures (ICLR 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.
