Deep Learning for Science

Over the past decade, Deep Learning powered algorithms have produced state-of-the-art results in Computer Vision, Speech Recognition and Robotics. Commercial Datacenters and HPC facilities now feature novel hardware to support AI workloads at scale. The international scientific community has deliberated around whether the third wave of neural networks is relevant for science. Multiple international reports have now been published, making a case for national investments in AI, and 
AI for Scientific applications.

This talk will delve into the frontier of Deep Learning for Scientific applications. Featuring use cases from Climate Science, Cosmology and High-Energy Physics, we’ll cover the broad landscape of analytics, simulation and control problems in science; and comment on applicability of modern Deep Learning architectures. The talk will conclude with a list of short and long term challenges for researchers interested in advancing this exciting frontier.