689917
event
1776791351
1776791366
<![CDATA[Ph.D. Proposal Oral Exam - Mohamed Mejri]]>
Title: Data and Energy-Efficient Learning Representations and Vector Symbolic Processing for Reasoning and Decision-MakingCommittee:
Dr. Chatterjee, Advisor
Dr. Hao, Chair
Dr. Mukhopadhyay
]]>
-
-
The objective of this research is to develop data- and energy-efficient artificial intelligence architectures that enable robust reasoning and real-time decision-making in resource-constrained environments. By leveraging Vector Symbolic Architectures (VSAs) and brain-inspired learning frameworks, the proposed system overcomes the massive data and power requirements of traditional large models to facilitate abstract reasoning and continual online learning. Additionally, it incorporates dynamic computational adaptability and high error resilience for harsh edge deployments, with future applications targeting complex robotic systems that unify perception, reasoning, and actuation.]]>
-
-
-
-
-
-
-
-
<![CDATA[]]>
-
-
- 434371
-
1788
-
102851
-
1808