Research map / Optimization / Optimizers
End-to-end Stochastic Optimization with Energy-based Model
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 |
|---|---|---|---|
| train the EBM with CD-LD instead of importance sampling (ablation) | SO-EBM · Power generator scheduling | task loss: 3.83 -> 3.85 | No clear effect |
| train the EBM with CD-LD instead of importance sampling (ablation) | SO-EBM · Power generator scheduling | training time per epoch (s): 0.68 -> 1.47 | Hurt |
| use SO-EBM instead of the two-stage predict-then-optimize model | load forecasting NN · Power generator scheduling | task loss: 4.52 -> 3.83 | Helped |
| use SO-EBM instead of DFL-QPTH | load forecasting NN · Power generator scheduling | task loss: 7.3% | Helped |
| use SO-EBM instead of DFL-QPTH | load forecasting NN · Power generator scheduling | training time per epoch (s): 93.12 -> 0.68 | Helped |
| use SO-EBM instead of DFL | GRU forecaster · COVID-19 resource allocation | task loss | No clear effect |
| use SO-EBM instead of DFL | GRU forecaster · COVID-19 resource allocation | training time: 4.3 times faster | Helped |
| use SO-EBM or DFL instead of the two-stage model | GRU forecaster · COVID-19 resource allocation | task loss: 10.9%+ | Helped |
| use SO-EBM instead of DFL-Block | attacker model · Network security game | task loss | Helped |
| use SO-EBM instead of Surrogate | attacker model · Network security game | task loss | No clear effect |
| use SO-EBM instead of Surrogate | attacker model · Network security game | training time | Hurt |
| use Policy-Net instead of the two-stage model | load forecasting · Power generator scheduling | task loss: 2x-3x larger | Hurt |
Closest papers by meaning
- Solver-Free Decision-Focused Learning for Linear Optimization Problems (NeurIPS 2025 · similarity 0.87)
- Joint Learning of Energy-based Models and their Partition Function (ICML 2025 · similarity 0.85)
- Feasibility-Aware Decision-Focused Learning for Predicting Parameters in the Constraints (NeurIPS 2025 · similarity 0.84)
- DFF: Decision-Focused Fine-Tuning for Smarter Predict-Then-Optimize with Limited Data (AAAI 2025 · similarity 0.84)
- BPQP: A Differentiable Convex Optimization Framework for Efficient End-to-End Learning (NeurIPS 2024 · similarity 0.83)
- Differentiable Distributionally Robust Optimization Layers (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.