DURIP GPU Computing for Enabling Causal Reasoning and Explainable Decision Making on Complex Networks
Abstract
Agency Program Manager: Dr. Predrag NeskovicEmail: predrag.neskovic.civ@us.navy.milProgram ManagerDivision 311, Office of Naval ResearchRequested Equipment. This project requests graphics processing unit (GPU) computing equipment,including a high-performance GPU server and a data server, to support the proposed researchand educational activities. The GPU server owns two Intel Xeon Platinum 8460Y processors andeight A100 GPUs (80GB), and the data server owns two Intel Xeon Gold 6326 Processors and264TB hard drives. The requested equipment offers an extremely useful computing resource thatwill have a lasting impact on research and education.Research Objectives. The requested high-performance GPU computing equipment will supporta research project on modeling and understanding of complex interventions on large and complexnetworks, which has been proposed to ONR recently. The key idea is to investigate theoreticalfoundations of causal inference on networks and develop new computational tools for modeling,monitoring, interpreting, and evaluatingcomplex interventions on large-scale complex networks.As considerable computational costs are expected in the proposed research, the requested GPUserver will facilitate the proposed research activities and accelerate the research progress. Moreover,the requested GPU server and data server will support the PI and collaborators at the Universityof Virginia (UVA) to explore sever emerging research topics on explainable causal learning,open-set recognition on networks, model calibration of graph neural networks, graph continuallearning, and large foundation models for networks. The PI is willing to leverage the proposedequipment to produce more preliminary results for future DOD white papers and full proposals. Insummary, the requested GPU server will support the proposed research project, help establish newresearch capabilities on causal reasoning and explainable decision making, and contribute to theeffective implementation and timely success of the research milestones.Educational Benefits. Extensive educational activities at UVA will be benefited from the requestedGPU computing equipment. Over 300 students at the School of Data Science and Departmentof Computer Science at UVA are taking courses on machine learning and artificial intelligenceevery year. Many of them are willing to engage to the DOD-relevant research topics on deeplearning, causal inference, network science, decision making, etc. The requested GPU computingequipment will provide critical computing resources to these students through course projects, facultymentored research for undergraduate students, and graduate research projects. Along with theproposed research activities, the GPU server will also bring DOD-relevant research opportunitiesto students who are the next generation of scientists and engineers.This abstract is publicly releasable.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Feb 24, 2025
- Source ID
- N000142512141
Entities
People
- Sheng Li
Organizations
- Office of Naval Research
- United States Navy
- University of Virginia