7 papers · 1 filter
Decentralized Parameter-Free Online Learning with Compressed Gossip
Tomas Ortega, Hamid Jafarkhani
We study decentralized online convex optimization when agents communicate over a graph and messages may be compressed. Classical decentralized online methods typically require lear…
Communication Compression for Distributed Learning with Aggregate and Server-Guided Feedback
Tomas Ortega, Chun-Yin Huang, Xiaoxiao Li +1
Distributed learning, particularly Federated Learning (FL), faces a significant bottleneck in the communication cost, particularly the uplink transmission of client-to-server updat…
Decentralized Parameter-Free Online Learning
Tomas Ortega, Hamid Jafarkhani
We propose the first parameter-free decentralized online learning algorithms with network regret guarantees, which achieve sublinear regret without requiring hyperparameter tuning.…
Communication Compression for Distributed Learning without Control Variates
Tomas Ortega, Chun-Yin Huang, Xiaoxiao Li +1
Distributed learning algorithms, such as the ones employed in Federated Learning (FL), require communication compression to reduce the cost of client uploads. The compression metho…
Offline Stochastic Optimization of Black-Box Objective Functions
Juncheng Dong, Zihao Wu, Hamid Jafarkhani +2
Many challenges in science and engineering, such as drug discovery and communication network design, involve optimizing complex and expensive black-box functions across vast search…
Decentralized Optimization in Time-Varying Networks with Arbitrary Delays
Tomas Ortega, Hamid Jafarkhani
We consider a decentralized optimization problem for networks affected by communication delays. Examples of such networks include collaborative machine learning, sensor networks, a…