What type of study is this?

This is a Quantitative Study study.

October 20, 2025Open Access

Efficient Online Mirror Descent Stochastic Approximation for Multi-Stage Stochastic Programming

View Full Paper

Ask AI

Bookmark

Authors

This analysis reveals the oracle complexity reduction in stochastic programming, highlighting the role of mirror descent methods.

The proposed method reduces oracle complexity from exponential to linear in the number of stages through semi-online decision making.
Using accelerated mirror descent algorithms, convergence is achieved both in expectation and with high probability.
The concept of stochastic conditional gradient oracles offers a novel multi-stage approach to stochastic programming.
Building decision variables several stages in advance allows asynchronous implementation of mirror descent algorithms.

Zhang et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark