Skip to content
Branch: master
Find file History
Permalink
Type Name Latest commit message Commit time
..
Failed to load latest commit information.
cifar10 small comment fix Dec 12, 2019
cifar100
imagenet Add MNIST experiments Nov 8, 2019
mnist Add MNIST experiments Nov 8, 2019
README.md change we to to Nov 21, 2019
classifier_comparison.ipynb Update examples for version 0.0.9 api Jul 28, 2019
classifier_comparison.py Update examples for version 0.0.9 api Jul 28, 2019
drawing_polyplices.ipynb
drawing_polyplices.py Changed package name from confidentlearning to cleanlab Nov 21, 2018
iris_simple_example.ipynb remove references to rankpruning Nov 8, 2019
iris_simple_example.py Update iris_simple_example.py Nov 8, 2019
model_selection_demo.ipynb Update examples for version 0.0.9 api Jul 28, 2019
model_selection_demo.py Update examples for version 0.0.9 api Jul 28, 2019
simplifying_confident_learning_tutorial.ipynb remove unused unicode future import Aug 20, 2019
simplifying_confident_learning_tutorial.py remove unused unicode future import Aug 20, 2019
visualizing_confident_learning.ipynb
visualizing_confident_learning.py Update examples for version 0.0.9 api Jul 28, 2019

README.md

cleanlab Examples

Not sure where to start? Try checking out how to find ImageNet Label Errors.

A brief description of the files and folders:

  • imagenet, 'cifar10', 'mnist' - code to find label errors in these datasets and reproduce the results in the confident learning paper. You will also need to git clone confidentlearning-reproduce.
  • classifier_comparison.ipynb - tutorial showing cleanlab performance across 10 classifiers and 4 dataset distributions.
  • iris_simple_example.ipynb - tutorial showing how to use cleanlab on the simple IRIS dataset.
  • model_selection_demo.ipynb - tutorial showing model selection on the cleanlab's parameter settings.
  • simplifying_confident_learning_tutorial.ipynb - tutorial implementing cleanlab as raw numpy code.
  • visualizing_confident_learning.ipynb - tutorial to demonstrate the noise matrix estimation performed by cleanlab.
You can’t perform that action at this time.