Research map / Learning theory and algorithms / Combinatorial and search optimization
A Novel Approach for Constrained Optimization in Graphical Models
Sara Rouhani, Tahrima Rahman, Vibhav Gogate
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 small k (3, 5) instead of large k (7, 9) in the k-separator (ablation) | ANYTIME-CMPE · Easy CMPE problems (DBN_16, Grids_11, Segmentation_12) | error (q - objective) | Improved |
| use k = 1 instead of k = 3, 5 (ablation) | ANYTIME-CMPE · Easy CMPE problems | error | Hurt |
| use k = 7, 9 instead of k = 3, 5 (ablation) | ANYTIME-CMPE · Easy CMPE problems | error: 1-2 orders of magnitude worse | Hurt |
| use large k (7, 9) instead of small k (ablation) | ANYTIME-CMPE · Hard CMPE problems (Grids_12/13/14/16/17/18, Segmentation_12/14/15) | error | Improved |
| use k = 1 instead of larger k (ablation) | ANYTIME-CMPE · Hard CMPE problems | error | Hurt |
| increase running time (100 to 1100 s) (ablation) | ANYTIME-CMPE with large k · Easy and hard CMPE problems | error | Improved |
| increase running time (100 to 1100 s) (ablation) | ANYTIME-CMPE with small k · Easy and hard CMPE problems | error | No clear effect |
| use ANYTIME-CMPE instead of the SCIP MILP encoding | CMPE solving · Easy CMPE problems (DBN_16, Grids_11, Segmentation_12), 1200 s | error | Improved |
| use SCIP instead of Gurobi as MILP baseline | MILP baseline · CMPE benchmarks | performance | Improved |
Closest papers by meaning
- Novel Upper Bounds for the Constrained Most Probable Explanation Task (NeurIPS 2021 · similarity 0.87)
- Bayesian Optimization of Functions over Node Subsets in Graphs (NeurIPS 2024 · similarity 0.86)
- How Do You Want Your Greedy: Simultaneous or Repeated? (JMLR 2023 · similarity 0.86)
- Counting the Optimal Solutions in Graphical Models (NeurIPS 2019 · similarity 0.86)
- Sparse Logistic Regression Learns All Discrete Pairwise Graphical Models (NeurIPS 2019 · similarity 0.85)
- Constrained Bayesian Optimization under Partial Observations: Balanced Improvements and Provable Convergence (AAAI 2024 · similarity 0.84)
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.
