Fairness in Machine Learning - Science in the News

It's no secret that bias is present everywhere in our society, from our educational institutions to the criminal justice system. The manifestation of this bias can be as seemingly trivial as the timing of a judge's lunch break or, more often, as fraught as race or economic class. We tend to attribute such discrimination to our own internalized prejudices and our inability to make decisions in truly objective ways. Because of this, machine learning algorithms seem like a compelling solution: we can write software to look at the data, crunch the numbers, and tell us what decision we should make.