Research map / Learning theory and algorithms / Combinatorial and search optimization
MAP Inference Via $$\ell _2$$-Sphere Linear Program Reformulation
Baoyuan Wu, Li Shen, Tong Zhang, Bernard Ghanem
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 |
|---|---|---|---|
| add the l2-sphere constraint to the LP relaxation (LS-LP) instead of plain LP relaxation methods | LP-relaxation MAP inference · Grids | solution type: LP-LP, PSDD and AD3 produce uniform solutions on all models; LS-LP gives valid solutions | Improved |
| use LS-LP instead of l2-Box ADMM | MAP inference · Seg-21 | mean logPot: -344.51 -> -324.89 | Improved |
| use LS-LP instead of l2-Box ADMM | MAP inference · Grids M14 | logPot: 44397.5 -> 48766.1 | Improved |
| use LS-LP instead of MPLP-C | MAP inference · Grids M14 | logPot: 45813.6 -> 48766.1 | Improved |
| use LS-LP instead of MPLP-C | MAP inference · Grids M1 | logPot: 3015.7 -> 2931.8 | Hurt |
| use LS-LP instead of MPLP-C | MAP inference · Protein M1 | logPot: -30181.3 -> -30165.5 | Improved |
| use LS-LP instead of ogm-ICM | MAP inference · Protein M5 | logPot: -256489 -> -267814 | Hurt |
| use LS-LP instead of LP-LP (active-set LP solver) | MAP inference · Seg-21 | runtime (s): 45.71 -> 342 | Hurt |
| use ogm-ICM instead of LP-LP | MAP inference · Seg-2 | mean logPot: -75.5 -> -137.1 | Hurt |
| add a perturbation epsilon to the objective and constraint of the ADMM (ablation) | LS-LP · Convergence analysis | global convergence guarantee | Improved |
Closest papers by meaning
- Exact MAP-Inference by Confining Combinatorial Search With LP Relaxation (AAAI 2018 · similarity 0.91)
- Exact MAP Inference by Avoiding Fractional Vertices (ICML 2017 · similarity 0.90)
- Partial Optimality by Pruning for MAP-Inference with General Graphical Models (TPAMI 2016 · similarity 0.90)
- Maximum Persistency via Iterative Relaxed Inference in Graphical Models (TPAMI 2017 · similarity 0.89)
- Continuous Relaxation of MAP Inference: A Nonconvex Perspective (CVPR 2018 · similarity 0.89)
- Convergence of Some Convex Message Passing Algorithms to a Fixed Point (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.
