Metadata-Version: 2.1
Name: moval
Version: 0.1.1
Summary: Model evaluation without manual labels
Home-page: https://github.com/ZerojumpLine/MOVAL
Author: Zeju Li
Author-email: lizeju8@gmail.com
Project-URL: Bug Tracker, https://github.com/AdaptiveMotorControlLab/MOVAL/issues
Classifier: Development Status :: 3 - Alpha
Classifier: Environment :: GPU :: NVIDIA CUDA
Classifier: Intended Audience :: Science/Research
Classifier: Operating System :: OS Independent
Classifier: Programming Language :: Python :: 3
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Classifier: License :: Free for non-commercial use
Requires-Python: >=3.8
Description-Content-Type: text/markdown
Requires-Dist: scikit-learn>=1.3.0
Requires-Dist: scipy>=1.8.0
Requires-Dist: pytest
Requires-Dist: gdown
Requires-Dist: pandas
Requires-Dist: nibabel

# MOVAL


A python package to evaluate model performance without the ground truth label.

## User Document

The latest documentation can be found [here](https://moval.readthedocs.io/en/latest/index.html).

## Reference

```
@article{li2022estimating,
  title={Estimating Model Performance under Domain Shifts with Class-Specific Confidence Scores},
  author={Li, Zeju and Kamnitsas, Konstantinos and Islam, Mobarakol and Chen, Chen and Glocker, Ben},
  journal={arXiv preprint arXiv:2207.09957},
  year={2022}
}
```
