presc 0.1.2

Performance Robustness Evaluation for Statistical Classifiers

PRESC: Performance and Robustness Evaluation for Statistical Classifiers

PRESC is a toolkit for the evaluation of machine learning classification models. Its goal is to provide insights into model performance which extend beyond standard scalar accuracy-based measures and into areas which tend to be overlooked in application, including:

• Generalizability of the model to unseen data for which the training set may not be representative
• Sensitivity to statistical error and methodological choices
• Performance evaluation localized to meaningful subsets of the feature space
• In-depth analysis of misclassifications and their distribution in the feature space

More details about the specific features we are considering are presented in the project roadmap. We believe that these evaluations are essential for developing confidence in the selection and tuning of machine learning models intended to address user needs, and are important prerequisites towards building trustworthy AI.

As a tool, PRESC is intended for use by ML engineers to assist in the development and updating of models. It will be usable in the following ways:

• As a standalone tool which produces a graphical report evaluating a given model and dataset
• As a Python package/API which can be integrated into an existing pipeline
• As a step in a Continuous Integration workflow: evaluations run as a part of CI, for example, on regular model updates, and fail if metrics produce unacceptable values.

We are using the standard Python scientific stack (numpy/pandas/jupyter). In order to streamline development while the project is still in its early stages, we are restricting focus to scikit-learn supervised classification models, and we are prototyping report visualizations in Jupyter notebooks. For the time being, the following are considered out of scope:

• Models built in machine learning frameworks other than scikit-learn
• User-facing evaluations, eg. explanations
• Evaluations which depend explicitly on domain context or value judgements of features, eg. protected demographic attributes. A domain expert could use PRESC to study misclassifications across such protected groups, say, but the PRESC evaluations themselves should be agnostic to such determinations.
• Analyses which do not involve the model, eg. class imbalance in the training data

There is a considerable body of recent academic research addressing these topics, as well as a number of open-source projects solving related problems. Where possible, we plan to offer integration with existing tools which align with our vision and goals.

This project was the subject of an Outreachy internship during Summer 2020. Submissions from the Spring 2020 application period have been archived in this this repo in their original state, and will be integrated here as needed.