2022/09/28 by Sai Aparna Aketi, Aketi, Sai Aparna, Sangamesh Kodge +3
Computer Science · Social Sciences · #Age of Information Optimization #Distributed #FOS: Computer and information sciences #Human Mobility and Location-Based Analysis #Machine Learning (cs.LG) #Multiagent Systems (cs.MA) #Parallel #Privacy-Preserving Technologies in Data #and Cluster Computing (cs.DC)
paper · pdf · doi:10.48550/arxiv.2209.14390
openalex publication_date 2022/09/28 · openalex created_date 2022/10/01 · openalex updated_date 2026/07/28
Decentralized learning over distributed datasets can have significantly different data distributions across the agents. The current state-of-the-art decentralized algorithms mostly assume the data distributions to be Independent and Identically Distributed. This paper focuses on improving decentralized learning over non-IID data. We propose Neighborhood Gradient Clustering (NGC), a novel decentralized learning algorithm that modifies the local gradients of each agent using self- and cross-gradient information. Cross-gradients for a pair of neighboring agents are the derivatives of the model parameters of an agent with respect to the dataset of the other agent. In particular, the proposed method replaces the local gradients of the model with the weighted mean of the self-gradients, model-variant cross-gradients (derivatives of the neighbors' parameters with respect to the local dataset), and data-variant cross-gradients (derivatives of the local model with respect to its neighbors' datasets). The data-variant cross-gradients are aggregated through an additional communication round without breaking the privacy constraints. Further, we present CompNGC, a compressed version of NGC that reduces the communication overhead by 32 ×. We theoretically analyze the convergence rate of the proposed algorithm and demonstrate its efficiency over non-IID data sampled from various vision and language datasets trained. Our experiments demonstrate that NGC and CompNGC outperform (by 0-6%) the existing SoTA decentralized learning algorithm over non-IID data with significantly less compute and memory requirements. Further, our experiments show that the model-variant cross-gradient information available locally at each agent can improve the performance over non-IID data by 1-35% without additional communication cost.