Exponentially-shifted Gaussian smoothing yields zeroth-order gradient estimators with linear dimension dependence, enabling improved complexity bounds for stochastic optimization including decision-dependent regimes.
Decision-dependent stochastic optimization: The role of distribution dynamics
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
years
2026 3representative citing papers
A projected primal-dual online algorithm is shown to track performatively stable saddle points with bounded mean-square error that decomposes into stochasticity, measurement, time-variation, and dual-set mismatch terms.
Introduces a two-stage robust optimization model with decision-dependent uncertainty sets to capture evolving manipulation costs and reduce gaming in strategic classification.
citing papers explorer
-
Complexity Guarantees for Zeroth-order Methods via Exponentially-shifted Gaussian Smoothing: Mitigating Dimension-dependence and Incorporating Decision-dependence
Exponentially-shifted Gaussian smoothing yields zeroth-order gradient estimators with linear dimension dependence, enabling improved complexity bounds for stochastic optimization including decision-dependent regimes.
-
Online Feedback Optimization for Constrained Stochastic Problems with Decision-Dependent Distributions: Extended Version
A projected primal-dual online algorithm is shown to track performatively stable saddle points with bounded mean-square error that decomposes into stochasticity, measurement, time-variation, and dual-set mismatch terms.
-
Robust Strategic Classification under Decision-Dependent Cost Uncertainty
Introduces a two-stage robust optimization model with decision-dependent uncertainty sets to capture evolving manipulation costs and reduce gaming in strategic classification.