Research map / Label-efficient and robust learning / Domain adaptation and generalization
Evidential Turing Processes
Melih Kandemir, Abdullah Akgül, Manuel Haussmann, Gozde Unal
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 changed | Where | Measured | Effect |
|---|---|---|---|
| Use ETP instead of BNN | IMDB sentiment classification · IMDB | negative test log-likelihood: 0.47 -> 0.37 | Improved |
| Use ETP instead of ENP | Fashion classification · Fashion | ECE (%): 3.7 -> 2.6 | Improved |
| Use ETP instead of ENP | SVHN classification · SVHN | negative test log-likelihood: 0.33 -> 0.26 | Improved |
| Use ETP instead of ENP | CIFAR10 classification · CIFAR10 | test error (%): 14.8 -> 15.3 | Hurt |
| Use ETP instead of EDL | CIFAR100 classification · CIFAR100 | ECE (%): 5.3 -> 6.6 | Hurt |
| Use ETP instead of ENP | MNIST out-of-domain detection for Fashion · MNIST | AUROC (%): 88.9 -> 90.0 | Improved |
| Use ETP instead of ENP | CIFAR100 out-of-domain detection for SVHN · CIFAR100 | AUROC (%): 92.4 -> 90.0 | Hurt |
| Use ETP instead of BNN | Random-input out-of-domain detection for IMDB · Random | AUROC (%): 60.9 -> 59.1 | Hurt |
| Use ETP instead of ENP | TinyImageNet out-of-domain detection for CIFAR100 · TinyImageNet | AUROC (%): 100.0 -> 99.6 | Hurt |
| Use ETP instead of competing models across the five classification tasks | Total calibration · Five real-world classification tasks | joint performance across test error, ECE, negative test log-likelihood, and OOD | Mixed |
Closest papers by meaning
- Uncertainty Estimation by Flexible Evidential Deep Learning (NeurIPS 2025 · similarity 0.89)
- Are Uncertainty Quantification Capabilities of Evidential Deep Learning a Mirage? (NeurIPS 2024 · similarity 0.88)
- The Unreasonable Effectiveness of Deep Evidential Regression (AAAI 2023 · similarity 0.88)
- Stop Guessing: Choosing the Optimization-Consistent Uncertainty Measurement for Evidential Deep Learning (ICLR 2026 · similarity 0.88)
- Is Epistemic Uncertainty Faithfully Represented by Evidential Deep Learning Methods? (ICML 2024 · similarity 0.87)
- Epistemic Uncertainty Quantification For Pre-Trained Neural Networks (CVPR 2024 · similarity 0.86)
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.
