AI Paper Atlas

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

Data Pruning via Moving-one-Sample-out

Haoru Tan, Sitong Wu, Fei Du, Yukang Chen, Zhibin Wang, Fan Wang, XIAOJUAN QI

NeurIPS 2023 · Learning theory and algorithms › Prediction and decision losses · approach: Distributionally robust optimization · 6 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
partition the training set into 10 subsets instead of 1 (ablation)MoSo · CIFAR-100 pruning ratio 0.2Top-1 accuracy: 74.35 -> 75.81Improved
partition the training set into 5 subsets instead of 1 (ablation)MoSo · CIFAR-100 pruning ratio 0.2Top-1 accuracy: 74.35 -> 75.76Improved
partition into 10 subsets instead of 5 (ablation)MoSo · CIFAR-100 pruning ratio 0.2Top-1 accuracy: 75.76 -> 75.81No clear effect
train surrogate for 200 epochs instead of 50 (ablation)MoSo · CIFAR-100 pruning ratio 0.2Top-1 accuracy: 75.76 -> 76.58No clear effect
train surrogate for 150 epochs instead of 100 (ablation)MoSo · CIFAR-100 pruning ratio 0.2Top-1 accuracy: 76.41 -> 76.19No clear effect
remove training-dynamics awareness (use last-epoch gradient norm only) (ablation)MoSo · CIFAR-100 across pruning ratesTop-1 accuracyHurt
lower the time step sampling rate (ablation)MoSo · CIFAR-100 pruning ratio 0.4Top-1 accuracyHurt
use MoSo instead of SOTA baselines at high pruning ratiosMoSo · CIFAR-100 and Tiny-ImageNetTop-1 accuracyImproved
use EL2N instead of random selection at high pruning ratiosEL2N · CIFAR-100Top-1 accuracyHurt
MoSo versus Herding under 20% label noiseMoSo · CIFAR-100 and Tiny-ImageNet noisy labelsTop-1 accuracy: less than 1%No clear effect

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