<div dir="auto">Dear Autonians, <div dir="auto"><br></div><div dir="auto">Amanda is a friend of the Lab and a coorganizer of our series of NeurIPS workshops on ML for the Developing World (ML4D).</div><div dir="auto"><br></div><div dir="auto">She will be defending her thesis this Thursday. </div><div dir="auto"><br></div><div dir="auto">The topic is relevant to at least a few of us.</div><div dir="auto"><br></div><div dir="auto">Cheers,</div><div dir="auto">Artur</div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">---------- Forwarded message ---------<br>From: <strong class="gmail_sendername" dir="auto">Amanda Coston</strong> <span dir="auto"><<a href="mailto:acoston@andrew.cmu.edu">acoston@andrew.cmu.edu</a>></span><br>Date: Mon, Apr 24, 2023, 11:01 AM<br>Subject: Thesis defense on principled machine learning for high-stakes decisions<br>To: <br></div><br><br><div dir="ltr">Hi!<div><br></div><div>Would you be able to attend my thesis (on Zoom) this Thursday at 3:30 pm? Details below.</div><div><br></div><div>Thanks in advance!</div><div>Amanda<br><div><br></div><div><div><b>Location</b></div><div>I will be defending virtually over zoom.</div><div>Topic: Thesis Defense - Amanda Coston<br>Time: Apr 27, 2023 03:30 PM Eastern Time (US and Canada)<br><br>Zoom Meeting<br><a href="https://cmu.zoom.us/j/91007768295?pwd=WkRoWmxQckR6MEE0U0Z2YWVuM2loQT09" target="_blank" rel="noreferrer">https://cmu.zoom.us/j/91007768295?pwd=WkRoWmxQckR6MEE0U0Z2YWVuM2loQT09</a><br><br>Meeting ID: 910 0776 8295<br>Passcode: 813156<br></div><div><br></div><div><b>Title: </b>Principled machine learning for societally consequential decision making</div><div><b><br></b></div><div><b>Abstract: </b> Machine learning algorithms are widely used for decision-making in societally high-stakes settings such as child welfare, criminal justice, healthcare, hiring, and consumer lending. Recent history has illuminated numerous examples where these algorithms proved unreliable or inequitable. This thesis proposes a principled approach to the use of machine learning in societally high-stakes settings, guided by three pillars: validity, equity, and oversight. We address data problems that challenge the validity of algorithmic decision support systems by developing methods for algorithmic risk assessments that account for selection bias, confounding, and bandit feedback. A central focus of this research is identifying when algorithms and human decisions disproportionately impact marginalized groups. Throughout we propose novel methods that use doubly-robust techniques for bias correction. We present empirical analysis in the child welfare, criminal justice and consumer lending domains.</div><div><b><br></b></div><div><b>Committee: </b>Edward Kennedy, Alexandra Chouldechova, Hoda Heidari, & Sendhil Mullainathan (U Chicago).</div><div><br></div><div dir="ltr" data-smartmail="gmail_signature"><div dir="ltr"><div>Regards,</div><div><br></div><div>Amanda Coston </div><div>PhD candidate in Machine Learning and Public Policy</div><div><a href="http://www.cs.cmu.edu/~acoston/" target="_blank" rel="noreferrer">amandacoston.com</a></div></div></div></div></div></div>
</div>