aamini/evidential-deep-learning
A Python library implementing evidential deep learning for training neural networks to estimate their own uncertainty directly from data.
Not currently ranked — collecting fresh signals.
star history
This repository provides TensorFlow/Keras and PyTorch implementations of evidential layers and loss functions based on a NeurIPS 2020 paper. The library enables neural networks to learn uncertainty estimates through evidential prior distributions (Normal Inverse-Gamma), adding evidential layers such as DenseNormalGamma to existing models. It supports both regression and classification tasks, providing calibrated confidence measures alongside predictions.
Frequently asked
- What is aamini/evidential-deep-learning?
- A Python library implementing evidential deep learning for training neural networks to estimate their own uncertainty directly from data.
- Is evidential-deep-learning open source?
- Yes — aamini/evidential-deep-learning is open source, released under the Apache-2.0 license.
- What language is evidential-deep-learning written in?
- aamini/evidential-deep-learning is primarily written in Python.
- How popular is evidential-deep-learning?
- aamini/evidential-deep-learning has 526 stars on GitHub.
- Where can I find evidential-deep-learning?
- aamini/evidential-deep-learning is on GitHub at https://github.com/aamini/evidential-deep-learning.