alibi
alibi copied to clipboard
Algorithms for explaining machine learning models
Updates the requirements on [numba](https://github.com/numba/numba) to permit the latest version. Release notes Sourced from numba's releases. Version 0.56.0 This release continues to add new features, bug fixes and stability improvements...
Hi, I was wondering if it is possible to plot a Confidence Interval around the ALE plots? I'm looking for something similar to the confidence interval implemented in PyALE.
I have trained an Autoencoder classifier and I am trying to get SHAP computations to run on multiple CPUs. However, I get an exception as follows - > TypeError: can't...
In `AnchorImage` the channel dimension is assumed to be the last dimension as defined [here](https://github.com/SeldonIO/alibi/blob/ab7dbf9b2a153ff49f29a372d5bdeac56dc7284a/alibi/explainers/anchor_image.py#L235). This is not necessarily true, for example in mnist handwritten pytorch model the shape of...
* Addresses issue https://github.com/SeldonIO/alibi/issues/687 . * Changed storage file format from joblib to tar.gz notes: In order to load .npy files from a tar archive, the changed function uses the...
Bump transformers to 4.16.0 in the future to be able to use `numpy` arrays in [Integrated Gradients transformers example](https://github.com/SeldonIO/alibi/blob/master/doc/source/examples/integrated_gradients_transformers.ipynb).
[sklearn.preprocessing.OneHotEncoder](https://scikit-learn.org/stable/modules/classes.html#module-sklearn.preprocessing) exposes multiple parameters such as: `drop`, `handle_unknown`, etc. which are useful to avoid overparametrisation (e.g. overparametrisation of binary variables, dummy trap for linear regression etc.) or handle unknown values....
Interventional TreeSHAP works properly with only up to 100 instances in the background dataset. This issue has been previously reported [here](https://github.com/slundberg/shap/issues/2487) and [here](https://github.com/slundberg/shap/issues/1991), mentioning what is the potential cause of...
The `test_save_cfrl` test in `alibi/tests/saving.py` has started failing intermittently. This seems to have started occurring as a result of a change in a dependency as the error could not be...
The `test_abdm_mvdm` infrequently fails due to the distance returned by one of `abdm` or `mvdm` being less than 0.