Discipline: Computer Sciences & Information Management
Subcategory: STEM Research
Lei Huang - Prairie View A&M University
The classical scientific computing method to simulate the physical phenomena of wave motion is to use either the finite difference or finite element methods to solve the partial differential equations representing the physical rules. Deep learning is a data-driven numerical optimization solution based on statistics and probabilities. Both solutions have pros and cons to solve scientific problems. In this poster, we present our research efforts to compare and converge these two solutions to simulate the seismic wave motions in geophysics. We will compare the performance and accuracy of these two solutions, and discuss how to converge them to achieve superior performance. This work is supported by NSF HBCU-UP/CNS #1832034.
Funder Acknowledgement(s): NSF HBCU-UP/CNS #1832034
Faculty Advisor: None Listed,
NSF Affiliation: HBCU-UP