AI Paper Atlas

Research map / Continual learning / Online and replay-based

Continual Learning by Asymmetric Loss Approximation With Single-Side Overestimation

Dongmin Park, Seokil Hong, Bohyung Han, Kyoung Mu Lee

ICCV 2019 · Continual learning › Online and replay-based · approach: Theoretical analysis of forgetting · 42 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 SI with ALASSO (asymmetric overestimated loss plus exact quadratic coefficient)SI · Permuted MNIST (30 tasks)accuracy A30: 0.752 -> 0.944Improved
replace SI with ALASSOSI · Split CIFAR-10/CIFAR-100accuracy A30: 0.632 -> 0.697Improved
replace SI with ALASSOSI · Split CIFAR-10/CIFAR-100forgetting F30: 0.110 -> 0.026Improved
replace SI with ALASSOSI · Split Tiny ImageNetaverage validation accuracy: approximately 10% point higherImproved
replace SI with ALASSOSI · Split CIFAR-10/CIFAR-100accuracy: 5.7% point better; 41% of lost accuracy recoveredImproved
replace SI with ALASSOSI · Permuted MNIST (30 tasks)intransigence I30: 0.008 -> 0.015Hurt
ALASSO versus single_task upper boundsingle_task · Split Tiny ImageNetaverage validation accuracy: 58.9 -> 59.4Improved
add parameter decoupling (ablation)ALASSO · Permuted MNISTaccuracy and convergence: slight accuracy gain, faster convergenceNo clear effect
add exact quadratic approximation of Omega (ablation)ALASSO · Permuted MNISTaccuracyImproved
add asymmetric overestimation of the unobserved side (ablation)ALASSO · Permuted MNISTaccuracyImproved
increase overestimation factor a from 1.0 (symmetric) to 2.0 (ablation)ALASSO · Permuted MNIST (100 tasks)accuracy: 0.62 -> 0.79Improved
increase a from 2.0 to 5.0 (ablation)ALASSO · Permuted MNIST (100 tasks)accuracy: 0.79 -> 0.75Hurt

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