AI Paper Atlas

Research map / Label-efficient and robust learning / Domain adaptation and generalization

Evidential Turing Processes

Melih Kandemir, Abdullah Akgül, Manuel Haussmann, Gozde Unal

ICLR 2022 · Label-efficient and robust learning › Domain adaptation and generalization · approach: Mixture and distillation for domain shift · 0 citations (OpenAlex)

Read the paper

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 ETP instead of BNNIMDB sentiment classification · IMDBnegative test log-likelihood: 0.47 -> 0.37Improved
Use ETP instead of ENPFashion classification · FashionECE (%): 3.7 -> 2.6Improved
Use ETP instead of ENPSVHN classification · SVHNnegative test log-likelihood: 0.33 -> 0.26Improved
Use ETP instead of ENPCIFAR10 classification · CIFAR10test error (%): 14.8 -> 15.3Hurt
Use ETP instead of EDLCIFAR100 classification · CIFAR100ECE (%): 5.3 -> 6.6Hurt
Use ETP instead of ENPMNIST out-of-domain detection for Fashion · MNISTAUROC (%): 88.9 -> 90.0Improved
Use ETP instead of ENPCIFAR100 out-of-domain detection for SVHN · CIFAR100AUROC (%): 92.4 -> 90.0Hurt
Use ETP instead of BNNRandom-input out-of-domain detection for IMDB · RandomAUROC (%): 60.9 -> 59.1Hurt
Use ETP instead of ENPTinyImageNet out-of-domain detection for CIFAR100 · TinyImageNetAUROC (%): 100.0 -> 99.6Hurt
Use ETP instead of competing models across the five classification tasksTotal calibration · Five real-world classification tasksjoint performance across test error, ECE, negative test log-likelihood, and OOD Mixed

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