2017 Data Science Day

2017 Data Science Day

Sponsored by

O.C. Tanner

O.C. Tanner

Agenda

11:30 AM - 1:00 PM Data Science Job Fair
1:00 PM - 1:10 PM Welcome: Data Science at Utah
1:10 PM - 2:00 PM Panel: Data Science in Industry
2:00 PM - 3:30 PM Posters and Demos
3:30 PM - 4:50 PM Data Science + X Talks
5:00 PM - 6:00 PM Keynote
6:00 PM - 6:15 PM Poster Awards !!

Posters and Demos

We welcome all students, staff, and faculty at the University of Utah to sign up below to present a poster or demo at the Utah Data Science Day 2017.

Sign up as a presenter

Sign-up Deadline: Tuesday. Jan 10, 2017

Recruiters

We welcome all companies both local and beyond to station a booth, at no cost, at the Utah Data Science Day 2017.

Sign up for a recruiting booth


Confirmed Participation

Currently, we have confirmed participation from the following companies:

Keynote

Edo Liberty

Speaker: Edo Liberty

Bio: Edo Liberty is a Principal Scientist at Amazon AWS Machine Learning group. Prior to joining Amazon this year, he was head of Yahoo’s Independent Research in New York where he focused on scalable machine learning and data mining for Yahoo critical applications. He received his B.Sc in Physics and Computer Science from Tel Aviv university and his Ph.D in Computer Science from Yale University. After that, he was a Post-Doctoral fellow at Yale in Program in Applied Mathematics. His personal research interests include fast dimensionality reduction, clustering, streaming and online algorithms, machine learning, and large scale numerical linear algebra. His research has garnered best papers at KDD 2013, TechPulse 2012, and SODA 2011.

Talk Title: Distributed Streaming Algorithms in Realtime Data Mining

Abstract: This talk will introduce the distributed streaming computational model. In this model, different parts of the data are streamed to different machines that cannot communicate with each other. Moreover, each machine examines the data stream once and operates with severe memory limitations. This is de facto the standard setting in large-scale IoT applications, information security tracking, and dynamic monitoring, just to name a few. Even though this computational model is challenging, some remarkable algorithmic results enable a wide set of capabilities.

In this talk, I will explain the distributed streaming setting and its limitations. Then, I will show how to approximate frequencies of items in streams and how this can be used for threading Mail by large e-mail providers. After that, I will show a very new result (FOCS 2016) which solves optimally the streaming quantiles problem. I will conclude with a short demo of an open source library for streaming algorithms that I contribute to.

Amazon’s AI group is looking for full-time scientists, engineers, and interns. Don’t be shy about leaving a copy of your resume with me at the end of the talk or sending it to me by email (libertye@amazon.com).


Data Science + X Talks


Panel on Data Science in Industry