Friday, January 24
Title: Statistics at Google Scale
Abstract: Google is a big statistics engine that collects, organizes, summarizes, and analyzes data to provide users with information anywhere and at any time. At its core lie measurement, experimentation, and learning followed by implementation and more measurement, experimentation, and learning -- almost all of which is automated and self-monitoring. This talk will show how data science, statistical principles, and huge amounts of data are combined to analyze data at Google scale.
Bio: Diane Lambert is a statistician who has made a career out of learning how to wrestle with, and sometimes tame, data. She started in academia (she was tenured by the statistics department at Carnegie Mellon) and then moved to Bell Labs, where she became a Bell Labs Fellow and department chair. She left for Google, where she is now a research scientist working on applications ranging from monitoring a gigantic robotic tape drive system to measuring ad effectiveness and understanding users' engagement with Google apps. Recently, she served on a National Academy of Sciences panel on massive data.