Research map / Learning theory and algorithms / Submodular and game algorithms
Robust Guarantees of Stochastic Greedy Algorithms
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 |
|---|---|---|---|
| Stochastic greedy (mean µ of selection distribution) instead of exact greedy under cardinality constraint with k ≥ 2/(ε²µ) | STOCHASTIC-GREEDY for monotone submodular maximization · cardinality constraint (theory) | approximation ratio w.p. 1−e^{-εµk/2} | No clear effect |
| Use modular function instead of general monotone submodular in stochastic greedy | STOCHASTIC-GREEDY · modular objective, cardinality constraint (theory) | approximation ratio | Helped |
| Stochastic greedy under intersection of P matroids (in expectation) | STOCHASTIC-MATROID-GREEDY · intersection of P matroids (theory) | expected approximation ratio | No clear effect |
| Stochastic greedy under a single matroid constraint instead of uniform cardinality constraint | STOCHASTIC-GREEDY · matroid constraints (theory) | high-probability approximation guarantee | Hurt |
| Stochastic local search instead of greedy (consistent or inconsistent noise) | stochastic local search for submodular maximization · probabilistic uncertainty, modular example (theory) | expected approximation ratio | Hurt |
| Higher variance of selection noise distribution (exponential, uniform[0,5], Gaussian vs uniform) | STOCHASTIC-GREEDY · coverage on ego-network (333 nodes, 5038 edges) | function value vs budget k: small differences | Hurt |
| Stochastic local search with noise distribution vs deterministic local search | local search · coverage on ego-network | function value vs budget k | Hurt |
Closest papers by meaning
- Instance Specific Approximations for Submodular Maximization (ICML 2021 · similarity 0.91)
- Guarantees for Greedy Maximization of Non-submodular Functions with Applications (ICML 2017 · similarity 0.91)
- A Unified Approach to Submodular Maximization Under Noise (NeurIPS 2025 · similarity 0.90)
- Gradient Methods for Online DR-Submodular Maximization with Stochastic Long-Term Constraints (NeurIPS 2024 · similarity 0.88)
- Nearly Minimax Optimal Submodular Maximization with Bandit Feedback (NeurIPS 2024 · similarity 0.88)
- Discretely beyond $1/e$: Guided Combinatorial Algortihms for Submodular Maximization (NeurIPS 2024 · similarity 0.88)
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.