AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

On the Robustness of Average Losses for Partial-Label Learning

Jiaqi Lv, Biao Liu, Lei Feng, Ning Xu, Miao Xu, Bo An, Gang Niu, Xin Geng, Masashi Sugiyama

TPAMI 2023 · Label-efficient and robust learning › Recognition and long-tail · approach: Partial and multi-label learning · 30 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 bounded MAE loss instead of unbounded CCEABS partial-label learning · CIFAR-10 ConvNet case 4test accuracy (%): 17.29 -> 51.96Improved
use bounded MAE loss instead of CCEABS · CIFAR-10 MLP case 1test accuracy (%): 22.47 -> 41.46Improved
use MAE instead of CCEABS · Fashion-MNIST MLP case 4test accuracy (%): 25.76 -> 82.41Improved
use MAE instead of CCEABS · MNIST Linear case 4test accuracy (%): 67.78 -> 90.37Improved
use GCE instead of MAEABS · CIFAR-10 ConvNet case 4test accuracy (%): 51.96 -> 43.83Hurt
use GCE instead of MAEABS · CIFAR-10 ConvNet case 3test accuracy (%): 68.88 -> 71.65Improved
use MAE instead of GCE on a linear modelABS · Fashion-MNIST Linear case 1test accuracy (%): 81.50 -> 82.50No clear effect
add robust warm start (MAE for 20 epochs) to PRODENPRODEN · CIFAR-10 ConvNet case 3test accuracy (%): 64.59 -> 66.86Improved
add robust warm start to PRODEN with early stoppingPRODEN · CIFAR-10 ConvNet case 3test accuracy (%): 73.05 -> 75.99Improved
add robust warm start to PRODENPRODEN · CIFAR-100 ConvNet case 3test accuracy (%): 34.78 -> 36.27Improved
add robust warm start and early stopping to PRODENPRODEN · CIFAR-100 ConvNet case 4test accuracy (%): 37.37 -> 47.86Improved
add early stopping to PRODENPRODEN · CIFAR-100 ConvNet case 4test accuracy (%): 37.37 -> 47.29Improved

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