AI Paper Atlas

Research map / Optimization / Nonconvex and smooth optimization

Debiasing Conditional Stochastic Optimization

Lie He, Shiva Kasiviswanathan

NeurIPS 2023 · Optimization › Nonconvex and smooth optimization · approach: Stochastic variance reduction · 0 citations (OpenAlex)

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 changedWhereMeasuredEffect
Add second-order stochastic extrapolation to BSGD gradient estimatesBSGD · CSOsample complexity: \cO(\varepsilon^{-6}) -> \cO(\varepsilon^{-4.5})Improved
Add second-order stochastic extrapolation to BSpiderBoost gradient estimatesBSpiderBoost · CSOsample complexity: \cO(\varepsilon^{-5}) -> \cO(\varepsilon^{-3.5})Improved
Use stochastic extrapolation with nested variance reductionFCCO · FCCO, n = \Omega(\varepsilon^{-2/3})sample complexity: \cO(n\varepsilon^{-3}) -> \cO(\max\{\sqrt{n}/\varepsilon^{2.5}, 1/(\sqrt{n}\varepsilon^{4})\})Improved
Use stochastic extrapolation with nested variance reductionFCCO · FCCO, n = \cO(\varepsilon^{-2/3})sample complexity: \cO(n\varepsilon^{-3})No clear effect
Apply stochastic extrapolationBSGD, BSpiderBoost, and NestedVR · Invariant logistic regression, CSOerrorImproved
Apply stochastic extrapolationE-NestedVR · Invariant logistic regression, FCCOerrorImproved
Apply stochastic extrapolationBSGD, BSpiderBoost, and NestedVR · Instrumental variable regressionconvergence speedImproved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy