Title: AI and Scientific Computing: Algorithmic Alloys for Forecasting and Control of Complex Systems
Abstract: Scientific Computing (SC) and Artificial Intelligence (AI) are key drivers in our quest to understand, predict and control complex systems in engineering and the natural world. In this talk, I explore the frontiers and synergies of these two paradigms. I introduce algorithmic alloys of SC and AI to (i) reduce the computational cost of large scale simulations through generative learning (ii) provide closures for reduced order models using reinforcement learning. I also present a critical comparison of physics informed machine learning and SC for modeling and controlling fluid flows in both simulations and experiments. I aim to showcase the unprecedented, rich spectrum of discovery methods now available to science and argue for their effective integration to address some of the most challenging problems of our times.