AI Paper Atlas

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

Learning with Unsure Responses

Kunihiro Takeoka, Yuyang Dong, Masafumi Oyamada

AAAI 2020 · Label-efficient and robust learning › Recognition and long-tail · approach: Group-robust classification · 10 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 unsure loss (absolute form) as regularizerlogistic regression · MNIST synthetic single annotatorF1: 0.85 -> 0.88Improved
add unsure loss (absolute form)logistic regression · MNIST synthetic single annotatorAUC: 0.76 -> 0.80Improved
add unsure loss (log form)SVM · MNIST synthetic single annotatorF1: 0.85 -> 0.84No clear effect
add unsure loss (absolute form)SVM · MNIST synthetic single annotatorF1: 0.85 -> 0.87Improved
add unsure loss (squared form) to personal classifiersPC · MNIST synthetic DS annotatorsF1: 0.82 ± 0.10 -> 0.93 ± 0.03Improved
add unsure loss (log form) to personal classifiersPC · MNIST synthetic PC annotatorsF1: 0.78 ± 0.20 -> 0.83 ± 0.13Improved
add unsure loss (abs form)logistic regression · DOG-WOLF single annotatorF1: 0.80±0.05 -> 0.85±0.04Improved
add unsure loss (log form)logistic regression · DOG-WOLF single annotatorAUC: 0.81±0.01 -> 0.81±0.01No clear effect
add unsure loss (log form)SVM · DOG-WOLF single annotatorF1: 0.74 ±0.06 -> 0.74 ±0.06No clear effect
add unsure loss to personal classifiersPC · DOG-WOLF five annotatorsF1: 0.60±0.08 -> 0.63±0.06Improved
add unsure loss to personal classifiersPC · DOG-WOLF five annotatorsAUC: 0.50±0.07 -> 0.51±0.06No clear effect
treat unsure instances as unlabeled (semi-supervised) rather than via unsure lossPC · MNIST synthetic DS annotatorsF1: 0.82 ± 0.10 -> 0.63 ± 0.07Hurt

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