Research map / Learning theory and algorithms / Combinatorial and search optimization
e-PAL: An Active Learning Approach to the Multi-Objective Optimization Problem
Marcela Zuluaga, Andreas Krause, Markus Püschel
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 e-PAL instead of ParEGO | multi-objective optimization · SNW, NoC, SW-LLVM | number of evaluations: 30% to 70% fewer evaluations | Improved |
| use e-PAL instead of PAL | Pareto front search · SNW, NoC, SW-LLVM | runtime (evaluation cost 0): up to 420x speedup | Improved |
| use e-PAL instead of PAL | Pareto front search · SNW, NoC, SW-LLVM | samples to reach desired accuracy: up to 2.9x fewer samples | Improved |
| increase eps from 1% to 30% | e-PAL · SNW, NoC, SW-LLVM | prediction error and evaluations: error below 0.7% with under 50 evaluations -> error below 7% with under 30 evaluations | Mixed |
| increase eps | e-PAL runtime speedup over PAL · SW-LLVM | speedup: 45x -> 420x | Improved |
| use e-PAL instead of PAL | Pareto front search · SW-LLVM | error vs evaluations trade-off: PAL wins for a small range | Hurt |
| set eps = 0 with theoretical beta_t | e-PAL · SNW, NoC, SW-LLVM | prediction error: error 0 with fewer than 115 evaluations | Improved |
| evaluation cost 30 minutes instead of free | e-PAL speedup over PAL · SNW, NoC, SW-LLVM | speedup: up to 420x -> almost 3x | Hurt |
Closest papers by meaning
- Preference-Guided Diffusion for Multi-Objective Offline Optimization (NeurIPS 2025 · similarity 0.86)
- Amortized Active Generation of Pareto Sets (NeurIPS 2025 · similarity 0.86)
- Multi-Objective Bayesian Optimization with Active Preference Learning (AAAI 2024 · similarity 0.86)
- Learning Configurations for Data-Driven Multi-Objective Optimization (ICML 2025 · similarity 0.85)
- ParetoFlow: Guided Flows in Multi-Objective Optimization (ICLR 2025 · similarity 0.85)
- Training Greedy Policy for Proposal Batch Selection in Expensive Multi-Objective Combinatorial Optimization (ICML 2024 · similarity 0.85)
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.
