A team ofÂ Oak Ridge National Laboratory researchers has receivedÂ a three-year, $2 million contract from the Energy Department toÂ study the potential use of machine learning tools in scientific data analysis.
ORNL said Friday it will explore deep learning methods to help scientists understand massive data sets through theÂ Advances in Machine Learning to Improve Scientific Discovery at Exascale and Beyond project.
“This understanding can help scientists build and support new scientific theories, and help to design better materials,” saidÂ Thomas Potok, leader of ORNL’s computational data analytics group.
Potok will carry out theÂ ASCEND project with fellow researchersÂ Robert Patton, Chris Symons, Steven Young and Catherine Schuman.
The teams plans to use theÂ Titan supercomputer at ORNL to test high-performance computing methods and build a deep learning network that will work to process and interpret data from multiple sources such as sensors.
A Battelle-University of Tennessee joint venture manages the laboratory for DOE.