search

Trusted-AI / AIF360

A comprehensive set of fairness metrics for datasets and machine learning models, explanations for these metrics, and algorithms to mitigate bias in datasets and models.
https://aif360.res.ibm.com/
Apache License 2.0
2.46k stars 840 forks source link

[WIP] Add regression metric #260

Closed SSaishruthi closed 1 year ago

SSaishruthi commented 3 years ago

This PR adds regression metric to the toolkit

SSaishruthi commented 3 years ago

@nrkarthikeyan This is a basic skeleton I was thinking. We have a bunch of methods to group datasets and get indexes. I did not use existing methods. Rather created a function explaining the functionality. Can you please take a look and provide suggestions on skeleton updates?

I will make necessary changes

nrkarthikeyan commented 3 years ago

@SSaishruthi - can you verify the DCO thing and sign off?

nrkarthikeyan commented 3 years ago

Can we do this directly on the sklearn version? In that case we can follow the pattern in https://github.com/Trusted-AI/AIF360/blob/master/aif360/sklearn/metrics/metrics.py

Is there any reason why the classical version is needed?

nrkarthikeyan commented 3 years ago

One place to start if you want to use the "classic" version is to start with the regression dataset class https://github.com/Trusted-AI/AIF360/blob/master/aif360/datasets/regression_dataset.py and also create a regression metric class similar to https://github.com/Trusted-AI/AIF360/blob/master/aif360/metrics/classification_metric.py

@hoffmansc any thoughts?