Research map / Learning theory and algorithms / Combinatorial and search optimization
Efficient Parameter Importance Analysis via Ablation with Surrogates
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| replace algorithm runs with random-forest EPM predictions in ablation analysis | ablation analysis · six AClib configuration benchmarks | analysis CPU time: 33 to 14 727 times faster than racing-ablation | Helped |
| use surrogate instead of racing on heterogeneous instance sets | ablation analysis · SPEAR-SWV | important parameter identification (TP/FP/FN): 0/2/1 -> 1/1/0 | Helped |
| use surrogate instead of racing | ablation analysis · CLASP-WS | important parameter identification (TP/FP/FN): 0/0/1 -> 0/5/1 | Hurt |
| use surrogate instead of racing | ablation analysis · LPG-SATELLITE | important parameter identification (TP/FP/FN): 2/0/0 -> 1/1/1 | Hurt |
| use surrogate instead of racing | ablation analysis · AClib benchmarks overall | ablation path quality | No clear effect |
| predict expected penalized running time (EPARX) instead of mean penalized running time (ablation) | EPM surrogate · illustrative example, kappa=300 | predicted PAR2: 250 -> 293.43 | Mixed |
Closest papers by meaning
- Practical Bayesian Algorithm Execution via Posterior Sampling (NeurIPS 2024 · similarity 0.82)
- SCALE: Selective Resource Allocation for Overcoming Performance Bottlenecks in Mathematical Test-time Scaling (AAAI 2026 · similarity 0.81)
- Better Understandings and Configurations in MaxSAT Stochastic Local Search Solvers via Anytime Performance Analysis (AAAI 2025 · similarity 0.80)
- Learning Surrogates for Offline Black-Box Optimization via Gradient Matching (ICML 2024 · similarity 0.80)
- Automatic Core-Guided Reformulation via Constraint Explanation and Condition Learning (AAAI 2024 · similarity 0.80)
- Optimal ablation for interpretability (NeurIPS 2024 · similarity 0.80)
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.