TY - GEN
T1 - Nonasymptotic convergence rates for cooperative learning over time-varying directed graphs
AU - Nedic, Angelia
AU - Olshevsky, Alex
AU - Uribe, Cesar A.
N1 - Publisher Copyright:
© 2015 American Automatic Control Council.
PY - 2015/7/28
Y1 - 2015/7/28
N2 - We study the problem of cooperative learning with a network of agents where some agents repeatedly access information about a random variable with unknown distribution. The group objective is to globally agree on a joint hypothesis (distribution) that best describes the observed data at all nodes. The agents interact with their neighbors in an unknown sequence of time-varying directed graphs. Following the pioneering work of Jadbabaie, Molavi, Sandroni, and Tahbaz-Salehi and others, we propose local learning dynamics which combine Bayesian updates at each node with a local aggregation rule of private agent signals. We show that these learning dynamics drive all agents to the set of hypotheses which best explain the data collected at all nodes as long as the sequence of interconnection graphs is uniformly strongly connected. Our main result establishes a non-asymptotic, explicit, geometric convergence rate for the learning dynamic.
AB - We study the problem of cooperative learning with a network of agents where some agents repeatedly access information about a random variable with unknown distribution. The group objective is to globally agree on a joint hypothesis (distribution) that best describes the observed data at all nodes. The agents interact with their neighbors in an unknown sequence of time-varying directed graphs. Following the pioneering work of Jadbabaie, Molavi, Sandroni, and Tahbaz-Salehi and others, we propose local learning dynamics which combine Bayesian updates at each node with a local aggregation rule of private agent signals. We show that these learning dynamics drive all agents to the set of hypotheses which best explain the data collected at all nodes as long as the sequence of interconnection graphs is uniformly strongly connected. Our main result establishes a non-asymptotic, explicit, geometric convergence rate for the learning dynamic.
UR - http://www.scopus.com/inward/record.url?scp=84940917154&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84940917154&partnerID=8YFLogxK
U2 - 10.1109/ACC.2015.7172262
DO - 10.1109/ACC.2015.7172262
M3 - Conference contribution
AN - SCOPUS:84940917154
T3 - Proceedings of the American Control Conference
SP - 5884
EP - 5889
BT - ACC 2015 - 2015 American Control Conference
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2015 American Control Conference, ACC 2015
Y2 - 1 July 2015 through 3 July 2015
ER -