Distributed Optimization Using the Primal-Dual Method of Multipliers
Title: Distributed Optimization Using the Primal-Dual Method of Multipliers
Recently, distributed signal processing has drawn increasing attention due to the demand for big-data processing and popularity of various types of networks (e.g., wireless sensor networks, smart grid, decentralized active noise control). The basic idea is to have a number of computing units (e.g., a sensor or a mobile phone) collaborate with neighboring ones in a distributed way to complete a complex task. Different types of methods have been proposed in the literature to enable distributed signal processing, such as belief propagation and distributed convex/nonconvex optimization. In this talk we present two distributed optimization algorithms, namely the alternating direction method of multipliers (ADMM) and the primal-dual method of multipliers (PDMM). Relationship between PDMM and Kalman filter will be discussed.
Dr. Guoqiang Zhang received the Bachelor degree from the University of Science and Technology of China (USTC) in 2003, a master of philosophy (M.Phil.) from the University of Hong Kong in 2006, and his PhD degree from the Royal Institute of Technology – KTH in 2010. He then worked as a post-doctoral researcher at Delft University of Technology full time until the end of 2014 and part time until the end of 2016. From 2015 to 2016, he worked as a senior researcher at Ercisson AB, Sweden. He is now working as a senior lecturer at the University of Technology Sydney. His current research interests include large scale distributed optimization, distributed control and signal processing in networks, and optimization in machine learning.