Research map / Optimization / Nonconvex and smooth optimization
Stochastic-Constrained Stochastic Optimization with Markovian Data
Yeongjong Kim, Dabeen Lee
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 ergodic DPP with adaptive time-varying parameters (EDPP-t) instead of vanilla DPP-t on Markovian data | drift-plus-penalty · Fairness-constrained logistic regression, Markov data | final constraint 1 infeasibility: 0.0800 -> -0.0904 | Improved |
| use adaptive parameters V_t, alpha_t (EDPP-t) instead of fixed horizon-based parameters (EDPP-T) (ablation) | EDPP · Fairness-constrained logistic regression, Markov data | final constraint 1 infeasibility: 0.2042 -> -0.0904 | Improved |
| use fixed horizon-based parameters (EDPP-T) instead of adaptive ones (ablation) | EDPP · Fairness-constrained logistic regression, Markov data | optimality gap after 20,000 samples: EDPP-T achieves the smallest optimality gap, followed by EDPP-t and MDPP | Improved |
| use fixed parameters (EDPP-T) instead of adaptive ones (ablation) | EDPP · Online convex optimization with ergodic constraints | cumulative constraint 1 violation: EDPP-T incurs a positive cumulative constraint 1 violation, while the others are negative | Hurt |
| use ergodic DPP (EDPP-t, EDPP-T, MDPP) instead of i.i.d. methods PD, PD2, DPP | constrained stochastic optimization · Fairness-constrained logistic regression, Markov data | optimality gap | Improved |
| use adaptive DPP-t instead of fixed-parameter DPP-T (ablation) | vanilla DPP · Fairness-constrained logistic regression, Markov data | optimality gap: DPP-t performs better than DPP-T | Improved |
| use EDPP-t on online convex optimization with ergodic constraints | EDPP-t · Online convex optimization with ergodic constraints | regret | Improved |
Closest papers by meaning
- Adapting to Mixing Time in Stochastic Optimization with Markovian Data (ICML 2022 · similarity 0.93)
- On Markov Chain Gradient Descent (NeurIPS 2018 · similarity 0.91)
- Improved Lower Bounds for First-order Stochastic Non-convex Optimization under Markov Sampling (ICML 2025 · similarity 0.90)
- Stochastic Optimization with Arbitrary Recurrent Data Sampling (ICML 2024 · similarity 0.89)
- Stochastic Gradients under Nuisances (NeurIPS 2025 · similarity 0.89)
- An Online Adaptive Sampling Algorithm for Stochastic Difference-of-convex Optimization with Time-varying Distributions (ICML 2025 · 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.
