Speaker: Dr. Alon Amit
Probability and statistics have a delightful capacity to confound and confuse. After demonstrating this capacity with a few toy examples, we will present several impossible statistical situations that actually occurred - in sports statistics, class-action lawsuits and various situations at some well-known software companies. Building on those examples we will try to demonstrate the importance - and subtlety - of properly choosing and interpreting metrics, both for machine learning and for human consumption.
Alon Amit (https://www.facebook.com/alon.amit) received a Ph.D. in Mathematics from the Hebrew University of Jerusalem, Israel, where he also obtained an M.Sc. in Mathematics and a B.Sc. in Mathematics, Computer Science and Physics. He is a Co-founder and VP of Product at Origami Logic (http://www.origamilogic.com/), a data exploration startup, and has previously worked as a Product Manager at Facebook and Google and as Vice President at Compugen, a biotechnology firm.
Alon is a frequent speaker at math (http://math.stanford.edu/circle/2013winter.php) circles (http://mathcircle.berkeley.edu/) around (http://www.marinmathcircle.org/) the San Francisco Bay Area, an officer of the Golden Section (http://sections.maa.org/golden/) of the MAA, and a contributor to Quora (http://www.quora.com/Alon-Amit), math.SE (http://math.stackexchange.com/users/308/alon-amit) and Mathoverflow (http://mathoverflow.net/users/25/alon-amit).