Recent Studies

Study shows how machine learning could improve COVID predictive models – Click here to read more.

New paper on Nature Physics Reviews: Physics-Informed Machine Learning

DeepMind Likes CRUNCH – see here their paper about our DeepOnet paper in Nature MI.

George Karniadakis received the 2021 SIAM/ACM Prize in Computational Science and Engineering for “advancing spectral elements, reduced-order modeling, uncertainty quantification, dissipative particle dynamics, fractional PDEs, and scientific machine learning, while pushing applications to extreme computational scales and mentoring many leaders.”

The Robot Scientists Are Coming. But that’s not a bad thing

This is an article by Jennifer Walter in Discover Magazine. This article explores the first machine to fully automate the scientific process and make a discovery of its own. These robot scientist combine artificial intelligence with robotic laboratory equipment. (Read article)

Extraction of mechanical properties of materials through deep learning from instrumented indentation

  George Karniadakis with his PhD senior student Lu Lu and collaborators from MIT and NTU (Singapore) have implemented a deep learning method named instrumented indentation as a means of extracting material properties from 3D printed materials. The properties of such materials (titanium and aluminum alloys) are very different than traditionally made materials, and it is difficult to infer them using existing methods. (Read more.)

Learning and Meta-Learning of Partial Differential Equations via Physics-Informed Neural Networks

Professor George Em Karnadakis from Brown University, in collaboration with Caltech, Stanford University, and the University of Utah, have been awarded an AFOSR MURI grant for their work in, “Learning and Meta-Learning of Partial Differential Equations via Physics-Informed Neural Networks:  Theory, Algorithms, and Applications.  This work seeks to overcome mathematical obstacles by introducing physics-informed learning, integrating seamlessly data and mathematical models, and implementing them using physics-informed neural networks (PINNS) and other new physics-informed networks (PINs).

In Silico Medicine Advances the Development of Sickle Cell Disease Therapies

Sickle Cell Disease affects approximately 100,000 people in the United States and millions worldwide.  African American experience this disease most prevalently, afflicting one out of every 365 babies.  New drug therapies are urgently needed, however the development of a novel drug requires sometimes four to six years of experimentation.  (Read full story.)

Mathematicians develop algorithm which can infer velocity and pressure from video of fluid flows

Professor George Karniadakis in collaboration with Maziar Raissi and Alireza Yazdani have developed an algorithm which could potentially be used to analyze magnetic resonance imagery of blood flow through a brain aneurysm and compute the stress being placed on an arterial wall.  Read full story and Science article.