Seid Koric, a MechSE research associate professor and a technical director at NCSA, presented aspects of his novel artificial intelligence research at the Data Science Institute’s Machine Learning for Industry (ML4I) forum on August 10.
The talk, “Confluence of Numerical Modeling Methods and Artificial Intelligence in Physics-based Simulations,” was co-presented with Dr. Shirui Luo from NCSA.
“State-of-the-art deep learning (DL) techniques require enormous datasets for successful training,” Koric and Luo stated in their abstract. “Physics-based FEA and CFD simulations using high-throughput and parallel capabilities of high-performance computing (HPC) are used to generate a large amount of training data for DL on thousands of simulated modeling scenarios. Once our innovative DL models are properly trained on HPC, they can instantly make inferences on any low-end computing system various forms of accurate modeling predictions when a novel input is presented. NCSA researchers have developed and used such data-driven or physics-informed surrogate deep learning models to accelerate modeling and design in topological optimization, highly nonlinear material responses, turbulence, and similarly computationally demanding workflows in science and engineering.”
The ML4I forum exists to foster the free exchange of ideas and best practices in applying ML/AI methods for solving practical industrial problems. Industrial participants work with the broader research community and solution vendors to identify the most promising solution strategies and demonstrate methods of data analytics that can achieve actionable outcomes. The industrial participants learn solution strategies for their own needs while the research community gains a broader perspective of ultimate use cases for ML/AI methods. The goal for ML4I is that all participants benefit by a tighter community bond between research and application of data analytical methods.