Event on 2013-05-04 17:30:00
David Larochelle and William Li will provide a high-level overview of machine learning by presenting a system that they’ve built to predict the authorship of U.S. Supreme Court decisions and using it to illustrate key concepts in supervised machine learning. The system will be described in detail as well as its application to cases in which the authorship is in doubt, including unsigned opinions and cases in which authorship may be in dispute, such as the 2012 Obamacare decision. Along the way, they will introduce key machine learning concepts such as cross validation, held out data, feature selection, and model comparison.
William Li is a graduate student at MIT in Electrical Engineering and Computer Science focusing on applied machine learning. He likes using real-world data to build practical applications, uncover insights, and answer questions. His work involves applying or developing tools, models, and algorithms in a variety of domains (including speech recognition, user activity prediction, assistive technology, natural language processing, and public policy) and communicating the results to diverse audiences. He has masters degrees from MIT in Computer science and the Technology and Policy Program and a bachelor’s degree biomedical engineering from University of Toronto.
David Larochelle is a developer at the Berkman Center for Internet & Society at Harvard where he focuses on using computational techniques to quantitatively understand online media. Prior to join Berkman he focused on information security. He is the co-author of Splint, an open source security vulnerability detection tool for C programs. He completed a Bachelor's degree in Economics and Computer Science at the College of William and Mary and a Master's in Computer Science and all but the proposal and dissertation for a Ph.D. in Computer Science at the University of Virginia.
Stay tuned for more details
at Artisan’s Asylum
10 Tyler Street
Somerville, United States