Active learning (machine learning)

Geodesy

Software engineering

Machine learning

Logistic regression

Odds

Benchmark (surveying)

Artificial intelligence

Geography

Labeled data

Computer science

Oracle

Classifier (UML)

arXiv (Cornell University)

Machine learning (ML) is increasingly being used in high-stakes applicationsimpacting society. Therefore, it is of critical importance that ML models donot propagate discrimination. Collecting accurate labeled data in societalapplications is challenging and costly. Active learning is a promising approachto build an accurate classifier by interactively querying an oracle within alabeling budget. We design algorithms for fair active learning that carefullyselects data points to be labeled so as to balance model accuracy and fairness.We demonstrate the effectiveness and efficiency of our proposed algorithms overwidely used benchmark datasets using demographic parity and equalized oddsnotions of fairness.

Fair Active Learning.