This repository has code for the paper High-Precision Model-Agnostic Explanations.
An anchor explanation is a rule that sufficiently “anchors” the prediction locally – such that changes to the rest of the feature values of the instance do not matter. In other words, for instances on which the anchor holds, the prediction is (almost) always the same.
At the moment, we support explaining individual predictions for text classifiers or classifiers that act on tables (numpy arrays of numerical or categorical data). If there is enough interest, I can include code and examples for images.
The anchor method is able to explain any black box classifier, with two or more classes. All we require is that the classifier implements a function that takes in raw text or a numpy array and outputs a prediction (integer)
The Anchor package is on pypi. Simply run:
pip install anchor_exp
Or clone the repository and run:
python setup.py install
See notebooks folder for tutorials.
- Tabular data
- Text data - see also this version on colab (thanks to Sam Havens), which downloads the data.
Here is the bibtex if you want to cite this work.