Research map / Optimization / Implicit bias and generalization
E2E-FS: An End-to-End Feature Selection Method for Neural Networks
Brais Cancela, Verónica Bolón‐Canedo, Amparo Alonso‐Betanzos
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 |
|---|---|---|---|
| use E2E-FS-Soft (moving M_rho) instead of E2E-FS | E2E-FS · CIFAR-100 (153 features) | accuracy: 40.48 -> 48.65 | Improved |
| use E2E-FS-Soft instead of E2E-FS | E2E-FS · Fashion-MNIST (39 features) | accuracy: 78.11 -> 86.16 | Improved |
| use E2E-FS-Soft instead of E2E-FS | E2E-FS · MNIST (39 features) | accuracy: 91.87 -> 95.44 | Improved |
| use E2E-FS-Soft instead of E2E-FS | E2E-FS · CIFAR-10 (1536 features) | accuracy: 93.01 -> 92.72 | Hurt |
| use E2E-FS-Soft instead of E2E-FS | E2E-FS · MNIST (392 features) | accuracy: 99.49 -> 99.30 | Hurt |
| use E2E-FS-Soft instead of DFS | WRN-16-4 · CIFAR-100 (153 features) | accuracy: 34.55 -> 48.65 | Improved |
| use E2E-FS-Soft instead of DFS | WRN-16-4 · MNIST (39 features) | accuracy: 95.73 -> 95.44 | No clear effect |
| use E2E-FS-Soft instead of iSFS | WRN-16-4 · CIFAR-10 (153 features) | accuracy: 64.15 -> 76.50 | Improved |
| use E2E-FS instead of SFS (training time) | WRN-16-4 · CIFAR-10 (153 features) | time: 7260s -> 4180s | Improved |
| use E2E-FS instead of Fisher | Naive network · GISETTE | AUC-BA: 0.918 -> 0.962 | Improved |
| use E2E-FS instead of Fisher | Naive network · LYMPHOMA | AUC-BA: 0.948 -> 0.933 | No clear effect |
| use E2E-FS instead of Fisher | Naive network · DEXTER | AUC-BA: 0.904 -> 0.924 | Improved |
Closest papers by meaning
- SAND: One-Shot Feature Selection with Additive Noise Distortion (ICML 2025 · similarity 0.87)
- Feature Selection using Stochastic Gates (ICML 2020 · similarity 0.87)
- A Probabilistic Derivation of LASSO and L12-Norm Feature Selections (AAAI 2019 · similarity 0.86)
- How does Gradient Descent Learn Features --- A Local Analysis for Regularized Two-Layer Neural Networks (NeurIPS 2024 · similarity 0.84)
- Indirectly Parameterized Concrete Autoencoders (ICML 2024 · similarity 0.84)
- Local Feature Selection without Label or Feature Leakage for Interpretable Machine Learning Predictions (ICML 2024 · similarity 0.83)
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.
