Alibi is an open source Python library aimed at machine learning model inspection and interpretation. The initial focus on the library is on black-box, instance based model explanations.
Goals¶
- Provide high quality reference implementations of black-box ML model explanation algorithms
- Define a consistent API for interpretable ML methods
- Support multiple use cases (e.g. tabular, text and image data classification, regression)
- Implement the latest model explanation, concept drift, algorithmic bias detection and other ML model monitoring and interpretation methods
Overview
Methods
Examples
- Anchor explanations for income prediction
- Anchor explanations on the Iris dataset
- Anchor explanations for movie sentiment
- Anchor explanations for ImageNet
- Anchor explanations for fashion MNIST
- Contrastive Explanations Method (CEM) applied to MNIST
- Contrastive Explanations Method (CEM) applied to Iris dataset
- Counterfactual instances on MNIST
- Counterfactuals guided by prototypes on MNIST
- Counterfactuals guided by prototypes on Boston housing dataset
- Trust Scores applied to Iris
- Trust Scores applied to MNIST
API reference