Research map / Learning theory and algorithms / Submodular and game algorithms
Approximate Decomposable Submodular Function Minimization for Cardinality-Based Components
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 |
|---|---|---|---|
| Use SparseCard with ε = 1.0 instead of the exact reduction | smallplant segmentation with 500 superpixels · smallplant, 500 superpixels | Sparsity: 0.013 | Helped |
| Use SparseCard with ε = 1.0 instead of the exact reduction | smallplant segmentation with 500 superpixels · smallplant, 500 superpixels | Runtime: 4.1 | Helped |
| Increase ε from 0.0002 to 1.0 (ablation) | SparseCard · smallplant, 500 superpixels | Runtime: 74.3 -> 4.1 | Helped |
| Increase ε from 0.0002 to 1.0 (ablation) | SparseCard · smallplant, 500 superpixels | Approximation factor minus 1: 7·10^{-7} -> 4·10^{-3} | Mixed |
| Use ε = 0.1 instead of ε = 1 for clique penalties (ablation) | SparseCard + HyperLocal · Stack Overflow local clustering, 45 clusters, clique penalty | F1: 0.56±0.19 -> 0.56±0.19 | No clear effect |
| Use ε = 0.1 instead of ε = 1 for clique penalties (ablation) | SparseCard + HyperLocal · Stack Overflow local clustering, 45 clusters, clique penalty | Relative runtime: 1.32±0.33 -> 1.81±0.43 | Hurt |
| Use ε = 0.1 instead of ε = 1 for x^0.9 penalties (ablation) | SparseCard + HyperLocal · Stack Overflow local clustering, 45 clusters, x^0.9 penalty | F1: 0.54±0.20 -> 0.54±0.21 | No clear effect |
| Use ε = 0.1 instead of ε = 1 for x^0.9 penalties (ablation) | SparseCard + HyperLocal · Stack Overflow local clustering, 45 clusters, x^0.9 penalty | Relative runtime: 1.17±0.25 -> 2.04±0.42 | Hurt |
| Use ε = 0.1 instead of ε = 1 for square-root penalties (ablation) | SparseCard + HyperLocal · Stack Overflow local clustering, 45 clusters, sqrt penalty | F1: 0.42±0.18 -> 0.42±0.19 | No clear effect |
| Use ε = 0.1 instead of ε = 1 for square-root penalties (ablation) | SparseCard + HyperLocal · Stack Overflow local clustering, 45 clusters, sqrt penalty | Relative runtime: 2.02±0.99 -> 3.19±1.4 | Hurt |
| Use approximate reductions for clique and x^0.9 cut penalties | HyperLocal · Stack Overflow local clustering | F1 | Helped |
| Use ε = 0.01 instead of larger approximation settings (ablation) | SparseCard + HyperLocal · Stack Overflow local clustering | F1 | No clear effect |
Closest papers by meaning
- Fast Decomposable Submodular Function Minimization using Constrained Total Variation (NeurIPS 2019 · similarity 0.90)
- Weakly Submodular Maximization Beyond Cardinality Constraints: Does Randomization Help Greedy? (ICML 2018 · similarity 0.89)
- Optimal approximation for unconstrained non-submodular minimization (ICML 2020 · similarity 0.89)
- Decomposable Submodular Function Minimization via Maximum Flow (ICML 2021 · similarity 0.89)
- Practical $0.385$-Approximation for Submodular Maximization Subject to a Cardinality Constraint (NeurIPS 2024 · similarity 0.87)
- Consistent Submodular Maximization (ICML 2024 · similarity 0.87)
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.