Announcements

• Colloquium on Digital Transformation Science

• 
  

  November 512, 3 pm CT Reconstructing SARS-COV-2 Response Pathways Mathematics of Deep Learning René Vidal, Herschel L. Seder Professor of Biomedical Engineering and Director of the Mathematical Institute for Data Science, Johns Hopkins Ziv Bar-Joseph, FORE Systems Professor of Computer Science, Carnegie Mellon University REGISTER FOR ZOOM WEBINAR SARS-CoV-2 is known to primarily impact cells via two viral entry factors, ACE2 and TMPRSS2. However, much less is currently known about virus activity within cells. We used computational methods based on probabilistic graphical models to integrate several recent SARS-CoV-2 interaction and expression datasets with general protein-protein and protein-DNA interaction datasets. The reconstructed models display the pathways viral proteins use to drive expression in human cells and the pathways the cell uses to respond to the infection. Intersecting key proteins on these pathways with expression data from underlying conditions shown to increase mortality from SARS-CoV-2, and with knockout and phosphorylation data, identifies a few potential targets for treating cells to reduce viral loads. The past few years have seen a dramatic increase in the performance of recognition systems, thanks to the introduction of deep networks for representation learning. However, the mathematical reasons for this success remain elusive. For example, a key issue is that the neural network training problem is non-convex, hence optimization algorithms may not return a global minima. In addition, the regularization properties of algorithms such as dropout remain poorly understood. The first part of this talk will overview recent work on the theory of deep learning that aims to understand how to design the network architecture, how to regularize the network weights, and how to guarantee global optimality. The second part of this talk will present sufficient conditions to guarantee that local minima are globally optimal and that a local descent strategy can reach a global minima from any initialization. Such conditions apply to problems in matrix factorization, tensor factorization, and deep learning. The third part of this talk will present an analysis of the optimization and regularization properties of dropout for matrix factorization in the case of matrix factorization. René Vidal is the Herschel Seder Professor of Biomedical Engineering and Director of the Mathematical Institute for Data Science at Johns Hopkins University. He is also an Amazon Scholar, Chief Scientist at NORCE, and Associate Editor in Chief of TPAMI. His current research focuses on the foundations of deep learning and its applications in computer vision and biomedical data science. He is an AIMBE Fellow, IEEE Fellow, IAPR Fellow, and Sloan Fellow, and has received numerous awards for his work, including the D’Alembert Faculty Award, J.K. Aggarwal Prize, ONR Young Investigator Award, NSF CAREER Award, and best paper awards in machine learning, computer vision, controls, and medical roboticsZiv Bar-Joseph is the FORE Systems Professor of Computational Biology and Machine Learning at Carnegie Mellon University. His work focuses on the development of machine learning methods for the analysis, modeling, and visualization of time series high throughput biological data. Dr. Bar-Joseph is the recipient of the Overton Prize, an NSF CAREER Award, and several conference Best Paper awards. He is currently leading the Computational Tools Center for the National Institutes of Health Human BioMolecular Atlas Program (HuBMAP). He has served on the advisory board of several national efforts including the National Institute for Allergy and Infectious Diseases Systems Biology Program. Software tools developed by his group are widely used for the analysis of genomics data.

  
  

Quick Links:

C3.ai DTI Webpage

Events

Information on Call for Proposals

Proposal Matchmaking

Training Materials (password protected)

C3 Administration (password protected)

Have Questions? Please contact one of us:

• Jay Roloff, jayr@illinois.edu (Executive Director, c3.ai.DTI)
• R. Srikant, rsrikant@illinois.edu (Co-Director, c3.ai.DTI)
• Tandy Warnow, warnow@illinois.edu (Co-chief Scientist, c3.ai.DTI)