<div dir="ltr">A relevant talk by an old friend of the Auton Lab.<div><br></div><div>Artur</div><div><br><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">---------- Forwarded message ---------<br>From: <strong class="gmail_sendername" dir="auto">Adam Perer</strong> <span dir="auto"><<a href="mailto:adamperer@cmu.edu">adamperer@cmu.edu</a>></span><br>Date: Mon, Oct 3, 2022 at 12:50 PM<br>Subject: [HCII Seminar] Rich Caruana - "Friends Don’t Let Friends Deploy Black-Box Models: The Importance of Intelligibility in Machine Learning"<br>To: <<a href="mailto:ml-all@cs.cmu.edu">ml-all@cs.cmu.edu</a>><br></div><br><br><div><div><div><div style="text-decoration-color:initial;text-decoration-style:initial;font-weight:400" dir="ltr" class="gmail_attr"><span>Rich Caruana, a </span>Senior Principal Researcher at Microsoft Research, will be giving a virtual HCII seminar <span>on Friday</span>, entitled, "<b><span><span> </span></span></b>Friends Don’t Let Friends Deploy Black-Box Models: The Importance of Intelligibility in Machine Learning".<br></div><div dir="ltr" style="font-weight:400;text-decoration-style:initial;text-decoration-color:initial"><div class="gmail_quote"><div dir="ltr"><div class="gmail_quote"><div dir="ltr"><div class="gmail_quote"><div dir="ltr"><div class="gmail_quote"><div dir="ltr"><div class="gmail_quote"><div dir="ltr"><div class="gmail_quote"><div dir="ltr"><div class="gmail_quote"><div dir="ltr"><div class="gmail_quote"><div dir="ltr"><div class="gmail_quote"><div dir="ltr"><div class="gmail_quote"><div lang="EN-US"><div><div><br></div><div>Details below:</div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div><div style="text-decoration-color:initial;text-decoration-style:initial;font-weight:400" dir="ltr" class="gmail_attr"><br></div><div style="text-decoration-color:initial;text-decoration-style:initial;font-weight:400" dir="ltr" class="gmail_attr"><b><span>Topic: HCII Fall Seminar Series</span></b><br></div><div style="text-decoration-color:initial;text-decoration-style:initial;font-weight:400" dir="ltr"><div class="gmail_quote"><div dir="ltr"><div class="gmail_quote"><div dir="ltr"><div class="gmail_quote"><div dir="ltr"><div class="gmail_quote"><div dir="ltr"><div class="gmail_quote"><div dir="ltr"><div class="gmail_quote"><div dir="ltr"><div class="gmail_quote"><div dir="ltr"><div class="gmail_quote"><div dir="ltr"><div class="gmail_quote"><div dir="ltr"><div class="gmail_quote"><div lang="EN-US"><div><p style="margin:0px" class="MsoNormal"><b><span>Time: <span> </span></span></b><b><span><span><span>Friday, October 7</span></span>, 2022<span> </span><span>- 1:30 PM Eastern Time (US and Canada) – Virtual </span></span></b><br></p><p style="margin:0px" class="MsoNormal"><span> </span><br></p><p style="margin:0px" class="MsoNormal"><span>The next presentation of the FALL 2022 Seminar session will be conducted on<span> </span><u>Friday, October 7, from 1:30 to 2:45 pm via the Zoom. </u>The featured speaker -- VIRTUALLY (REMOTELY) -- will be<span> </span><b>Rich Caruana from Microsoft Research. </b> </span><br></p><div><br></div><p style="margin:0px" class="MsoNormal"><b><span>Join Zoom Meeting:</span></b><br></p><p style="margin:0px" class="MsoNormal"><br></p><p style="margin:0px" class="MsoNormal"></p><div><a style="text-decoration-color:initial;text-decoration-style:initial;text-decoration-line:none;color:rgb(84,172,220)" rel="noopener noreferrer" href="https://cmu.zoom.us/j/98589029500?pwd=YzYrTkVUU1pvN1AwUXpkbU55SXpYZz09" target="_blank">https://cmu.zoom.us/j/98589029500?pwd=YzYrTkVUU1pvN1AwUXpkbU55SXpYZz09</a><br></div><div><br></div><div>Meeting ID: 985 8902 9500<br></div><div>Passcode: 916104<br></div><div><br></div><div><span><b>Rich Caruana, </b></span><b>Senior Principal Researcher at Microsoft Research in Redmond, WA</b><br></div><p></p><div><img src="cid:f295b9e8a6473090_0.1" alt="" width="207" height="207" style="max-width:100%;height:auto"><br></div><div><span><br></span><span><b>Pres</b></span><b>entation Title</b><b><span>: <span> </span></span></b> Friends Don’t Let Friends Deploy Black-Box Models: The Importance of Intelligibility in Machine Learning</div><p style="margin:0px" class="MsoNormal"><br></p><p style="margin:0px" class="MsoNormal"><b>Abstract<span style="color:rgb(31,73,125)">: </span></b><span> </span><span> </span><br></p><p style="margin:0px" class="MsoNormal">In machine learning, sometimes tradeoffs must be made between accuracy, privacy and intelligibility: the most accurate models usually are not very intelligible or private, and the most intelligible models usually are less accurate. This can limit the accuracy and privacy of models that can safely be deployed in mission-critical applications such as healthcare where being able to understand, validate, edit, and trust models is important. EBMs (Explainable Boosting Machines) are a recent learning method based on generalized additive models (GAMs) that are as accurate as full complexity models, more intelligible than linear models, and which can be made differentially private with little loss in accuracy. EBMs make it easy to understand what a model has learned and to edit the model when it learns inappropriate things. In the talk, I’ll present multiple case studies where EBMs discover surprising patterns in data that would have made deploying black-box models risky. I’ll describe how to train these glassbox models with boosted trees, and with deep neural nets, and I’ll briefly discuss how we’re using these models to uncover and mitigate bias in models where fairness and transparency are important. <br></p><p style="margin:0px" class="MsoNormal"><br></p><p style="margin:0px" class="MsoNormal"><b>Bio: </b><span style="color:rgb(31,73,125)"> </span><span> </span> <br></p><p style="margin:0px" class="MsoNormal"><span>Rich Caruana is a senior principal researcher at Microsoft Research. Before joining Microsoft, Rich was on the faculty in the Computer Science Department at Cornell University, at UCLA’s Medical School, and at CMU’s Center for Learning and Discovery. Rich’s Ph.D. is from Carnegie Mellon University, where he worked with Tom Mitchell and Herb Simon. His thesis on Multi-Task Learning helped create interest in a new subfield of machine learning called Transfer Learning. Rich received an NSF CAREER Award in 2004 for Meta Clustering, best paper awards in 2005 (with Alex Niculescu-Mizil), 2007 (with Daria Sorokina), and 2014 (with Todd Kulesza, Saleema Amershi, Danyel Fisher, and Denis Charles), and co-chaired KDD in 2007. His current research focus is on learning for medical decision making, transparent modeling, and deep learning.</span><br></p></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div><div><br></div></div><div></div><br><div><div><div dir="ltr"><div style="color:rgb(39,39,40);font-family:-apple-system,sans-serif;font-size:14px" dir="auto">Adam Perer ( <a style="color:var(--linkColor)" href="http://perer.org/" rel="noopener noreferrer" target="_blank">http://perer.org</a> )</div><div style="color:rgb(39,39,40);font-family:-apple-system,sans-serif;font-size:14px" dir="auto">Data Interaction Group ( <a style="color:var(--linkColor)" href="http://dig.cmu.edu/" rel="noopener noreferrer" target="_blank">http://dig.cmu.edu</a> )<br></div><div style="color:rgb(39,39,40);font-family:-apple-system,sans-serif;font-size:14px" dir="auto">Human-Computer Interaction Institute</div><div style="color:rgb(39,39,40);font-family:-apple-system,sans-serif;font-size:14px" dir="auto">Carnegie Mellon University</div></div></div><br></div></div></div>
</div></div></div>