Math Professor Receives $225,000 NSF Grant to Address Computational Arithmetic Issues

  • College of Arts & Sciences, Co-Director of Hive, A&S Creative & Technical Services
  • Director of Communications & Creative Services
  • Communications & Creative Services
  • Creative Content & Social Media
  • Hive
  • Integrated Strategic Communication
901 Patterson Office Tower
(859) 257-6819

By Jenny Wells

Qiang Ye, University of Kentucky professor of mathematics in the UK College of Arts and Sciences, has received funding from the National Science Foundation (NSF) to research and develop new algorithms for solving linear algebra problems that will address accuracy problems in computer arithmetic.
 
The three-year, $225,000 grant will allow Ye and his team to develop new methods to more accurately compute eigenvalues of large matrices, a computation that has many scientific and engineering applications such as Google search page ranking, structure design, image processing and circuit simulations.
 
Large-scale computations of this nature are often inherently ill-conditioned, according to Ye, which implies their results may suffer from loss of accuracy caused by round off errors. “The computational accuracy of numerical algorithms are difficult issues that are often overlooked,” Ye said. “By developing algorithms that can maintain accuracy as problem size increases, we improve reliability of these algorithms so that they can be more confidently used by practitioners.”
 
To learn more about the project, visit http://nsf.gov/awardsearch/showAward?AWD_ID=1620082.
 
 
UK is the University for Kentucky. At UK, we are educating more students, treating more patients with complex illnesses and conducting more research and service than at any time in our 150-year history. To read more about the UK story and how you can support continued investment in your university and the Commonwealth, go to: uky.edu/uk4ky. #uk4ky #seeblue
 
X
Enter your linkblue username.
Enter your linkblue password.
Secure Login

This login is SSL protected

Loading