Quick Links:

C3.ai DTI Webpage

Events

Information on Call for Proposals

Proposal Matchmaking (Archive from 2020 CFP)

C3.ai DTI Training Materials

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)

Announcing our Third Call for Proposals:

AI to Transform Cybersecurity and Secure Critical Infrastructure

It is anticipated that up to USD $10 million in Research Awards will be awarded from this Call for Proposals. Proposals can request funding of USD $100,000 to $1,000,000 for an initial period of one (1) year.

C3DTI will also make available up to USD $2 million in Azure Cloud computing resources, supercomputing resources at UIUC’s NCSA and LBNL’s NERSC, and free, unlimited access to the C3 AI Suite hosted on the Microsoft Azure Cloud. 

**** We will be offering a virtual session with information on the CFP including an opportunity to ask questions early in January. Please keep an eye out on you email and this wiki page for more information. ****

Proposals are Due February 7, 2022

Awards will be made in March 2022 with start dates of around June 1, 2022

Announcements

Colloquium on Digital Transformation Science

May 13, 3 pm CT

Graceful AI: Backward-Compatibility, Positive-Congruent Training, and the Search for Desirable Behavior of Deep Neural Networks

Vice President of Applied Science, Amazon Web Services, Professor of Computer Science, UCLA

REGISTER FOR ZOOM WEBINAR

As machine learning-based decision systems improve rapidly, we are discovering that it is no longer enough for them to perform well on their own. They should also behave nicely towards their predecessors and peers. More nuanced demands beyond accuracy now drive the learning process, including robustness, explainability, transparency, fairness, and now also compatibility and regression minimization. We call this “Graceful AI,'’ because in 2021, when we replace an old trained classifier with a new one, we should expect a peaceful transfer of decision powers. Today, a new model can introduce errors that the old model did not make, despite significantly improving average performance. Such “regression” can break post-processing pipelines, or cause the need to reprocess large amounts of data. How can we train machine learning models to not only minimize the average error, but also minimize “regression”? Can we design and train new learning-based models in a manner that is compatible with previous ones, so that it is not necessary to re-process any data? These problems are prototypical of the nascent field of cross-model compatibility in representation learning. I will describe the first approach to Backward-Compatible Training (BCT), introduced at the last Conference on Computer Vision and Pattern Recognition (CVPR), and an initial solution to the problem of Positive-Congruent Training (PC-Training), a first step towards “regression constrained” learning, to appear at the next CVPR. Along the way, I will also introduce methodological innovations that enable full-network fine-tuning by solving a linear-quadratic optimization. Such Linear-Quadratic Fine-Tuning (LQF, also to appear at the next CVPR) achieves performance equivalent to non-linear fine-tuning, and superior in the low-data regime, while allowing easy incorporation of convex constraints.

Quick Links:

C3.ai DTI Webpage

Events

Information on Call for Proposals

Proposal Matchmaking

C3.ai DTI Training Materials Overview (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)