AI Paper Atlas

Research map / Learning theory and algorithms / Prediction and decision losses

A nonparametric method for gradual change problems with statistical guarantees

Lizhen Nie, Dan L Nicolae

NeurIPS 2021 · Learning theory and algorithms › Prediction and decision losses · approach: Distributionally robust optimization · 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
use max-gap estimator ρ̌ instead of arg-max estimator ρ̂proposed method · Setting I quadratic locationaverage l1 localization error: 0.24 -> 0.15Improved
use max-gap estimator ρ̌ instead of arg-max estimator ρ̂proposed method · Setting I volatility linearaverage l1 localization error: 0.26 -> 0.15Improved
use max-gap estimator ρ̌ instead of arg-max estimator ρ̂proposed method · Setting I location linear d=50average l1 localization error: 0.09 -> 0.05Improved
use max-gap estimator ρ̌ instead of arg-max estimator ρ̂proposed method · Setting II network linearaverage l1 localization error: 0.07 -> 0.02Improved
use max-gap estimator ρ̌ instead of arg-max estimator ρ̂proposed method · Setting II location linear d=1average l1 localization error: 0.12 -> 0.14No clear effect
replace general-purpose ρ̂_gen with proposed ρ̌baseline ρ̂_gen · Setting I location linear d=1average l1 localization error: 0.17 -> 0.09Improved
replace abrupt-change KCpA with proposed ρ̌KCpA · Setting I location complex d=1average l1 localization error: 0.27 -> 0.03Improved
replace abrupt-change KCpA with proposed ρ̌KCpA · Setting II location complex d=1average l1 localization error: 0.29 -> 0.04Improved
replace abrupt-change KCpA with proposed ρ̌KCpA · Setting I network linearaverage l1 localization error: 0.16 -> 0.10Improved
use assumption-specific ρ̂_one-side instead of ρ̌ when the change is not one-sidedρ̂_one-side · Setting I location complex d=1average l1 localization error: 0.03 -> 0.62Hurt
use ρ̂_mix instead of ρ̌ as dimension growsρ̂_mix · Setting I location linear d=50average l1 localization error: 0.05 -> 0.43Hurt
use one-sided estimator ρ̂_one-side when its assumption holdsρ̂_one-side · Setting I location one-sided d=1average l1 localization error: 0.03 -> 0.02Mixed

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