Speaker Dr. Mark Embree

Professor Dr. Mark Embree, Virginia Tech, will give a presentation at Seton Hall University on March 24, 2007.

Two Challenges in Data Analytics: Educating Undergraduates and Learning from a Building

In 2015 Virginia Tech launched a new undergraduate major in Computational Modeling and Data Analytics.  The program, based on ten new courses jointly taught by the Computer Science, Mathematics, and Statistics Departments, has quickly attracted over 300 majors.  In the first part of the talk, we will describe this curriculum and share some lessons we have learned en route to graduating our first class in May.

The second part of the talk will focus on a particular data science application.  Virginia Tech’s Goodwin Hall is instrumented with 212 accelerometers welded to the steel frame of the building.  Each of these sensors records vibrations at the rate of up to 25,600 samples per second.  From this massive data set, the Virginia Tech Smart Infrastructure Lab seeks to learn about many aspects of building occupancy, ranging from day-to-day operating conditions to the detection of an emergency.  We will briefly describe the basics of interpolatory matrix factorizations (e.g., CUR decompositions) and show how they can be applied to identify a minimal set of sensors, toward establishing low-cost sensor configurations for instrumenting other buildings.  (This talk describes joint work with Pablo Tarazaga, Serkan Gugercin, students in the Virginia Tech Smart Infrastructure Lab, and colleagues in GAiTE LLC.)

Leave a Reply

Your email address will not be published. Required fields are marked *