2020 — 2023 |
Moss, J. Eliot Thomas, Philip |
N/AActivity Code Description: No activity code was retrieved: click on the grant title for more information |
Fmitf:Track I: Verified Safe and Fair Machine Learning @ University of Massachusetts Amherst
Artificial intelligence (AI), and specifically machine learning, is being used more and more in areas with significant real-world impacts on people's lives. Examples include the delivery of health care and social services, decisions in the legal-justice system, self-driving cars, and face and speech recognition. Researchers have discovered that these applications of machine learning often embody biases, or health, safety, or economic risks. This project's novelty lies in developing ways to show that a test of the safety or fairness of a machine-learning system is mathematically sound and correctly coded on a computer, so that its test results can be relied upon. The project's impacts will thus be greater assurance that risks (lack of safety) and biases (lack of fairness) are known and evaluated precisely and correctly.
The investigators develop computer-checked proofs of correctness of several components necessary to the overall goals described above. These computer-checked proofs of formulations of the necessary statistical tests, such as Hoeffding's Inequality (and other such inequalities), are used to bound the probability that bias or safety risk exceeds a given limit. The mathematics of these is known, but computer-checked proofs are novel. Further, some newer bounds have hand-written proofs possibly needing more rigor or stronger assumptions, the limitations of which will be revealed by attempting computer-checked proofs. Next, computer code used to implement the safety/fairness tests needs similar proofs of correctness. Some aspects of how to do this are well-known, but computer-checked proofs for the numerical (floating-point) computations involved are lacking, and challenging. Lastly, the researchers will improve the computer proof tools, which remain weak in certain respects, by using machine learning to assist in these kinds of proofs.
This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.
|
0.936 |
2022 — 2024 |
Khan, Javed [⬀] Thomas, Philip |
N/AActivity Code Description: No activity code was retrieved: click on the grant title for more information |
Cc* Compute: Accelerating Compute Driven Science Through a Sharable High Performance Computing Cluster in Kent State Multi-Campus System
This project will add an agile locally and globally sharable HPCC (High-Performance Computing Cluster) hosted in a ScienceDMZ enclave, integrated with national science computing facilities, including the Open Science Grid (OSG), by creatively using recent advances in federated science networking and distributed systems’ virtualization open to regional faculty. The system is composed of 18 nodes with dual Intel Xeon Gold 6242R class CPUs (20 core), 192GB RAM, and an NVIDIA A30 class GPUs. Storage is spread across the nodes using CEPH
The project supports several interesting newly emerging collaborative HPCC workflows- scienceware as-a-service (SAS) and science-data-lakes (SDL), and intense real-time-computing (iRTC) besides supporting the HPC and HTC workflows. NSF-funded resources in this project are open to all faculty researchers in northeast Ohio colleges and their collaborators, including the faculty of all eight campuses of Kent who are in the network’s latency proximity and engaged in data-intensive collaborative workflows. In order to support high throughput and collaborative computing, the ScienceDMZ exercises a new model of unimpeded host-centric cauterized and federated security, as opposed to the traditional perimeter focused security approach. It is already fronted by a 100-Gbps Data Transfer Node (DTN) capable of ‘friction-free’ long-haul transferring massive datasets.
The project directly contributes to NSF’s goals to foster innovation, integration, and engineering of new campus-level networking and cyberinfrastructure that can assertively support widely collaborative, multi-campus distributed massive-data driven research and harness largely untapped potential to share unused compute cycles and resources across the entire academic fabric, while leveraging a compelling set of science projects from a wide variety of disciplines.
This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.
|
0.948 |