What does it mean to talk about fairness within the context of machine learning algorithms, and how do we decide what is fair? Suchana Seth speaks about different definitions of fairness in this lightning talk.
Seth discusses various efforts by the tech industry to combat the pervasive problem of algorithmic bias and the ways in which technology can learn to compensate for the bias that it inherits from learned data. She argues that while there are various ways to make algorithms fairer, choosing the right definition of fairness is not as straightforward.
Data & Society’s Fellows Talks is a three-part Databite series showcasing our 2016-2017 fellows cohort. Each talk features 3 fellows speaking about their work, wide-ranging interdisciplinary connections, and a few of the provocative questions that have emerged this year.
Data & Society Executive Director Janet Haven moderates the conversation.
Suchana Seth is a physicist-turned-data scientist from India. She has built scalable data science solutions for startups and industry research labs, and holds patents in text mining and natural language processing. Suchana believes in the power of data to drive positive change, volunteers with DataKind, mentors data-for-good projects, and advises research on IoT ethics. She is also passionate about closing the gender gap in data science, and leads data science workshops with organizations like Women Who Code.
Data & Society’s “Databites” speaker series presents timely conversations about the purpose and power of technology, bridging our interdisciplinary research with broader public conversations about the societal implications of data and automation.