The paper investigates the effects of disruptions to peer-to-peer communications on decentralized learning in a disaster setting. The authors examine the resilience of a decentralized learning process when a subset of devices abruptly drops from the process. They found that the accuracy of the learning process is more affected by a loss of connectivity than by a loss of data. However, the network remains relatively robust, and the learning process can still achieve a good level of accuracy.
Publication date: 5 Oct 2023
Project Page: Not provided
Paper: https://arxiv.org/pdf/2310.02986
