Research map / Label-efficient and robust learning / Recognition and long-tail
Composite Feature Selection Using Deep Ensembles
Fergus Imrie, Alexander Luke Ian Norcliffe, Pietro Lio, Mihaela van der Schaar
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 |
|---|---|---|---|
| Discover groups with an ensemble of sparse feature-selection learners | CompFS · Syn1 | Gsim: 0.50 { $\pm$ 0.00} -> 0.91 { $\pm$ 0.14} | Improved |
| Discover groups with an ensemble of sparse feature-selection learners | CompFS · Syn2 | Gsim: 0.50 { $\pm$ 0.00} -> 0.90 { $\pm$ 0.20} | Improved |
| Discover groups with an ensemble of sparse feature-selection learners | CompFS · Syn3 | Gsim: 0.67 { $\pm$ 0.00} -> 0.68 { $\pm$ 0.05} | No clear effect |
| Discover groups with an ensemble of sparse feature-selection learners | CompFS · Syn4 | Gsim: 0.50 { $\pm$ 0.00} -> 0.47 { $\pm$ 0.20} | No clear effect |
| Discover groups with an ensemble of sparse feature-selection learners | CompFS · Chem1 | Gsim: 0.50 { $\pm$ 0.00} -> 0.82 { $\pm$ 0.20} | Improved |
| Discover groups with an ensemble of sparse feature-selection learners | CompFS · Chem2 | Gsim: 0.50 { $\pm$ 0.00} -> 0.72 { $\pm$ 0.24} | Improved |
| Discover groups with an ensemble of sparse feature-selection learners | CompFS · Chem3 | Gsim: 0.50 { $\pm$ 0.00} -> 0.62 { $\pm$ 0.17} | Improved |
| Use only 25 selected features rather than all 489 features | METABRIC prediction · METABRIC | AUROC: 0.869 -> 0.830 | Hurt |
| Use a single group instead of the multi-group ensemble (ablation) | CompFS · Synthetic and chemistry tasks | feature selection performance | Mixed |
| Increase the number of learners while keeping regularisation unchanged (ablation) | CompFS on Syn2 · Syn2 | number of discovered groups | No clear effect |
Closest papers by meaning
- Dual Feature Reduction for the Sparse-group Lasso and its Adaptive Variant (ICML 2025 · similarity 0.83)
- Enhancing Neural Subset Selection: Integrating Background Information into Set Representations (ICLR 2024 · similarity 0.82)
- Multi-group Learning for Hierarchical Groups (ICML 2024 · similarity 0.82)
- Estimating Conditional Mutual Information for Dynamic Feature Selection (ICLR 2024 · similarity 0.82)
- PEAKS: Selecting Key Training Examples Incrementally via Prediction Error Anchored by Kernel Similarity (ICML 2025 · similarity 0.82)
- SAND: One-Shot Feature Selection with Additive Noise Distortion (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.
