New UCB strategies enhance adaptive deep neural networks for edge computing

Researchers have introduced four new Upper Confidence Bound (UCB) strategies to Adaptive Deep Neural Networks (ADNNs) for edge computing environments. These strategies, including UCB-Bayes, UCB-Tuned, and UCB-V, aim to dynamically balance accuracy with energy consumption and latency. Experiments on ResNet and MobileViT models using CIFAR datasets showed that UCB-Bayes converged fastest, while UCB-V and UCB-Tuned offered the best trade-offs between accuracy, latency, and energy usage. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT Introduces new adaptive inference strategies for efficient deep learning on edge devices, potentially improving performance and resource utilization.

RANK_REASON This is a research paper introducing new algorithms and comparative analysis.

Read on arXiv cs.LG →

• Upper Confidence Bound
• Adaptive Deep Neural Networks
• ADNNs
• UCB1
• UCB-Bayes
• UCB-Tuned
• UCB-V
• ResNet
• MobileViT
• CIFAR-10
• CIFAR-100
• Grigorios Papanikolaou

• paper
• other