Session
Beyond Supervised Learning: Approaches for Efficient and Reliable Intelligence
Event TypeResearch Manuscript
Presented In-Person
AI/ML System Design
Design
TimeTuesday, December 7th2:35pm - 3:00pm PST
Location3018
DescriptionAs AI is entering into all aspects of our lives, issues such as scalability, privacy, and reliability become predominant in addition to performance. There exists an increasing interest in non-traditional methods to address the efficiency/complexity dilemma encountered in supervised learning. This session presents four diverse approaches along with federated learning, Hyper-dimensional computing, reinforcement learning, and verifiable deep learning.
Presentations
| 2:35pm - 3:00pm PST | A3C-S: Automated Agent Accelerator Co-Search towards Efficient Deep Reinforcement Learning |