Summary of Joint Model Pruning and Resource Allocation For Wireless Time-triggered Federated Learning, by Xinlu Zhang et al.
Joint Model Pruning and Resource Allocation for Wireless Time-triggered Federated Learning
by Xinlu Zhang, Yansha Deng, Toktam Mahmoodi
First submitted to arxiv on: 3 Aug 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Information Theory (cs.IT)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary A novel approach to federated learning is introduced, where users are organized into tiers based on fixed time intervals. The Time-triggered (TT) system faces challenges due to a growing number of devices and limited wireless bandwidth, leading to issues like stragglers and communication overhead. To mitigate these problems, model pruning is applied to TT systems, and the joint optimization problem of pruning ratio and bandwidth allocation is solved to minimize training loss under communication latency constraints. The proposed TT-Prune method demonstrates a 40% reduction in communication cost compared to asynchronous multi-tier FL without model pruning, while maintaining convergence at the same level. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Federated learning helps many devices learn together without sharing their data. A new way of organizing these devices into groups based on time is called Time-triggered (TT) federated learning. This approach has its own challenges, like when some devices take longer to send and receive information, causing delays. To fix this, researchers applied a technique called model pruning to reduce the amount of information being sent. They also solved an optimization problem to find the right balance between reducing the amount of data sent and meeting certain deadlines for completing training. This new approach, called TT-Prune, helps reduce communication costs by 40% while keeping the same level of accuracy. |
Keywords
» Artificial intelligence » Federated learning » Optimization » Pruning
