Research map / Learning theory and algorithms / Clustering
Contamination-source based K-sample clustering
Xavier Milhaud, Denys Pommeret, Yahia Salhi, Pierre Vandekerkhove
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 |
|---|---|---|---|
| add data-driven tuning of gamma and C (ablation) | KCMC k-sample test · Alternative, k=10 config C, n~400 | empirical power (%): 2 -> 63 | Improved |
| add data-driven tuning of gamma and C (ablation) | KCMC k-sample test · Alternative, k=10 config C, n~1000 | empirical power (%): 1 -> 99 | Improved |
| add data-driven tuning of gamma and C (ablation) | KCMC k-sample test · Alternative, k=7, n~1000 | empirical power (%): 1 -> 95 | Improved |
| add data-driven tuning of gamma and C (ablation) | KCMC k-sample test · Alternative, k=4, n~1000 | empirical power (%): 18 -> 50 | Improved |
| add data-driven tuning of gamma and C (ablation) | KCMC k-sample test · Alternative, k=4, n~3000 | empirical power (%): 12 -> 99 | Improved |
| increase sample size from about 400 to about 3000 | KCMC k-sample test (tuned) · Alternative, k=4 | empirical power (%): 20 -> 50 -> 99 | Improved |
| tune penalty and test with epsilon=0.87 fixed instead (ablation) | KCMC k-sample test · Null, empirical level | empirical level (%) | No clear effect |
| lower fixed penalty exponent epsilon to much smaller values | KCMC k-sample test · Null, finite samples | empirical level | Hurt |
| increase number of populations k from 2 to 10 | KCMC k-sample test · Null, n~400 | empirical level (%): 5 -> 5 | No clear effect |
| reduce n_i p_i below about 50 | KCMC k-sample test · Null level | empirical level | Hurt |
| increase sample size from about 300 to about 800 | KCMC clustering · Simulated clustering, K=10 | clustering accuracy: n≈300 -> n≈800 | Improved |
| replace KMBD k-means with KCMC at larger n | KCMC clustering · Simulated clustering, n≈1500 | clustering accuracy | Improved |
Closest papers by meaning
- Scaling Epidemic Inference on Contact Networks: Theory and Algorithms (NeurIPS 2025 · similarity 0.82)
- Aligning Time-series by Local Trends: Applications in Public Health (AAAI 2025 · similarity 0.80)
- Learning With Multi-Group Guarantees For Clusterable Subpopulations (ICML 2025 · similarity 0.80)
- Hierarchical and Density-based Causal Clustering (NeurIPS 2024 · similarity 0.80)
- Global-Local Dirichlet Processes for Clustering Grouped Data in the Presence of Group-Specific Idiosyncratic Variables (ICML 2025 · similarity 0.79)
- SPATIALEPIBENCH: Benchmarking Spatial Information and Epidemic Priors in Forecasting (NeurIPS 2026 · similarity 0.79)
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.
