Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Evidential Deep Learning to Quantify Classification Uncertainty
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Replace softmax probabilities with Dirichlet evidence outputs and train using the regularized expected L2 loss | LeNet · notMNIST | predictive entropy | Helped |
| Replace softmax probabilities with Dirichlet evidence outputs and train using the regularized expected L2 loss | LeNet · CIFAR10 last five categories | predictive uncertainty | Helped |
| Replace softmax probabilities with Dirichlet evidence outputs and train using the regularized expected L2 loss | LeNet · MNIST adversarial examples | uncertainty on wrong predictions | Helped |
| Replace softmax probabilities with Dirichlet evidence outputs and train using the regularized expected L2 loss | LeNet · CIFAR5 adversarial examples | uncertainty on wrong predictions | Helped |
| Use the expected L2 loss instead of Type II maximum likelihood or expected cross-entropy (ablation) | Evidential Deep Learning · classification | stability of uncertainty estimates | Helped |
| Add KL regularization toward the uniform Dirichlet for non-misleading evidence, with annealing (ablation) | Evidential Deep Learning · classification | uncertainty for samples the model cannot classify | Helped |
Closest papers by meaning
- Uncertainty Estimation by Flexible Evidential Deep Learning (NeurIPS 2025 · similarity 0.90)
- Hyper Evidential Deep Learning to Quantify Composite Classification Uncertainty (ICLR 2024 · similarity 0.89)
- Are Uncertainty Quantification Capabilities of Evidential Deep Learning a Mirage? (NeurIPS 2024 · similarity 0.88)
- Is Epistemic Uncertainty Faithfully Represented by Evidential Deep Learning Methods? (ICML 2024 · similarity 0.87)
- Uncertainty Regularized Evidential Regression (AAAI 2024 · similarity 0.87)
- Random-Set Neural Networks (ICLR 2025 · 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.