Individuals and communities increasingly coexist with machines—from traffic signals to pollution sensors to security alarms—that sense and share vast quantities of information collectively known as Big Sensory Data. These details are intended to make communities run more smoothly, but instead, they are overwhelming both the machines collecting the data and the people (like community leaders) analyzing it. Zhi Tian, a professor of electrical and computer engineering, is studying the collection and computation of BSD in hopes of making the solutions those data call for easier to bring to fruition.
For her study, Tian is examining five distinct problems: 1) data acquisition algorithms for physical world reconstruction; 2) data sampling algorithms that can guarantee arbitrary accuracy for various operations; 3) dominant data collection methods for retrieving primary components from BSD, and the corresponding primary component-based approximation algorithms; 4) integration model of multimodal BSD; and 5) approximate computation on low-quality BSD, considering the fact that the captured BSD may be inaccurate, incomplete, inconsistent, obsolete and affected by noises.
The ultimate outcomes of this research will include new methodologies, algorithms and tools—all of which will improve not just academic understanding, but real-world experiences for those who encounter smart technology.
Tian is the lead Principal Investigator of a collaborative team from four universities, including George Mason University, George Washington University, Georgia State University and Virginia Commonwealth University. The National Science Foundation awarded $1.2 millions to the team for this project, of which Tian received $282,528. She will begin her work in January 2018 and complete the project in late August 2020.
Article published on volgenau.gmu.edu