TY - GEN
T1 - Parallel multi-splitting proximal method for star networks
AU - Wei, Ermin
N1 - Publisher Copyright:
© 2017 American Automatic Control Council (AACC).
PY - 2017/6/29
Y1 - 2017/6/29
N2 - We develop a parallel algorithm based on proximal method to solve the problem of minimizing summation of convex (not necessarily smooth) functions over a star network. We show that this method converges to an optimal solution for any choice of constant stepsize for convex objective functions. Under further assumption of Lipschitz-gradient and strong convexity of objective functions, the method converges linearly.
AB - We develop a parallel algorithm based on proximal method to solve the problem of minimizing summation of convex (not necessarily smooth) functions over a star network. We show that this method converges to an optimal solution for any choice of constant stepsize for convex objective functions. Under further assumption of Lipschitz-gradient and strong convexity of objective functions, the method converges linearly.
UR - http://www.scopus.com/inward/record.url?scp=85027044357&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85027044357&partnerID=8YFLogxK
U2 - 10.23919/ACC.2017.7963623
DO - 10.23919/ACC.2017.7963623
M3 - Conference contribution
AN - SCOPUS:85027044357
T3 - Proceedings of the American Control Conference
SP - 4341
EP - 4346
BT - 2017 American Control Conference, ACC 2017
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2017 American Control Conference, ACC 2017
Y2 - 24 May 2017 through 26 May 2017
ER -