AI Paper Atlas

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

Moment Distributionally Robust Tree Structured Prediction

Yeshu Li, Danyal Saeed, Xinhua Zhang, Brian D Ziebart, Kevin Gimpel

NeurIPS 2022 · 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
replace BiAF cross-entropy layer with Marginal DRO loss, 10 training samplestree DRO · PTB m=10UAS: 93.48 -> 94.51Improved
replace BiAF layer with Marginal DRO loss, 10 samplestree DRO · UD Dutch m=10UAS: 90.86 -> 92.41Improved
replace BiAF layer with Marginal DRO loss, 10 samplestree DRO · CTB m=10UAS: 88.45 -> 89.19Improved
replace BiAF layer with Marginal DRO loss, 10 samples, low-resource Turkishtree DRO · UD Turkish m=10UAS: 17.64 -> 24.85Improved
replace BiAF layer with Marginal DRO loss with 1000 samplestree DRO · PTB m=1000UAS: 97.16 -> 97.12No clear effect
Game algorithm instead of BiAF with 1000 samplestree DRO Game · UD Turkish m=1000UAS: 42.82 -> 36.30Hurt
use Stochastic or Game instead of Marginal for training costtree DRO · CPU time per steptime (s): 0.28 -> 2.72Hurt
Game-theoretic solver instead of Marginaltree DRO · CPU time per steptime (s): 0.28 -> 7.25Hurt
use ADMM instead of Frank-Wolfe for arborescence projection (ablation)projection oracle · random-point projectionobjectiveImproved
use ADMM instead of Frank-Wolfe for projection (ablation)projection oracle · per-iteration costcost: about 8n times higherHurt
vary smoothing mu from tiny to moderate (ablation)Marginal · UD Dutch m=100UASImproved
small lambda (ambiguity radius) (ablation)Marginal · UD Dutch m=100UASHurt

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