Current browse context:
stat.ML
Change to browse by:
References & Citations
Mathematics > Numerical Analysis
Title: Constructing unbiased gradient estimators with finite variance for conditional stochastic optimization
(Submitted on 4 Jun 2022 (v1), last revised 13 Sep 2022 (this version, v3))
Abstract: We study stochastic gradient descent for solving conditional stochastic optimization problems, in which an objective to be minimized is given by a parametric nested expectation with an outer expectation taken with respect to one random variable and an inner conditional expectation with respect to the other random variable. The gradient of such a parametric nested expectation is again expressed as a nested expectation, which makes it hard for the standard nested Monte Carlo estimator to be unbiased. In this paper, we show under some conditions that a multilevel Monte Carlo gradient estimator is unbiased and has finite variance and finite expected computational cost, so that the standard theory from stochastic optimization for a parametric (non-nested) expectation directly applies. We also discuss a special case for which yet another unbiased gradient estimator with finite variance and cost can be constructed.
Submission history
From: Takashi Goda [view email][v1] Sat, 4 Jun 2022 13:28:44 GMT (105kb,D)
[v2] Fri, 26 Aug 2022 04:42:57 GMT (328kb,D)
[v3] Tue, 13 Sep 2022 00:02:05 GMT (328kb,D)
Link back to: arXiv, form interface, contact.