- This event has passed.
Data Science at Knewton
October 27, 2015 @ 6:30 pm - 10:00 pm
6:30 – 7:00 PM: Networking & Food
7:00 – 7:30 PM: Hilary’s talk
7:30 – 8:00 PM: Kevin’s talk
8:30 – 9:00 PM: Socializing
Adaptive learning technology is used to provide personalized learning experiences for students. A common implementation provides individualized follow-up assignments for struggling students. Using observed data, we show how to use bootstrapping and rejection sampling to create a valid comparative study of groups of students who do and don’t use adaptive technology, even when the samples are not uncorrelated.
Hillary Green-Lerman is a member of the research team at Knewton. As a data scientist at Knewton, she designs and implements studies that provide quantitative evidence of the potential of Knewton-powered products to improve student learning outcomes. Previously, Hillary worked as a scientific associate at D.E. Shaw Research, where she performed large-scale molecular dynamics simulations of protein-drug interactions on the Anton supercomputers, and published her findings in Nature. She holds a B.S. in Materials Science and Engineering from the University of California, Berkeley. Previously, she has spoken at NYC Python Meetup, Biophysical Society, and Protein Society, and has written about women in STEM for Lilith Magazine.
A common problem faced by the NBA and Knewton are evaluating (groups of) people’s skills. The NBA cares about who is the best basketball team. Knewton cares whether or not you have mastered a particular skill. These problems may be modeled in strikingly similar ways. In this talk, we step through why these situations may be modeled so similarly as well as how their different evaluation metrics affect our confidence in the models’ performance.
Kevin Wilson is a principal data scientist at Knewton. During his tenure there, he has helped build, implement, and evaluate the proficiency and other models which serve millions of students using Knewton’s recommendation engine. Before Knewton, he received his PhD in mathematics from Princeton and his bachelor’s in mathematics from the University of Michigan.