DocumentCode
696242
Title
Estimating the interconnection structure of dynamical networks
Author
Blackhall, Lachlan ; Rotkowitz, Michael
Author_Institution
Res. Sch. of Inf. Sci. & Eng., Australian Nat. Univ., Canberra, ACT, Australia
fYear
2009
fDate
23-26 Aug. 2009
Firstpage
2954
Lastpage
2959
Abstract
Complex dynamical networks emerge from the physical or information based interconnection of many dynamical systems. These networks display emergent behaviour that is best understood through knowledge of the interconnection structure of the network. We analyze and compare a variety of existing regression techniques (some sparsity inducing and other not) with a recursive sparse estimator, presented recently by the authors, for determining this interconnection structure. In large networks the ability to recursively estimate the interconnection structure of the network may be advantageous for a number of reasons and thus this work represents a proof-of-concept that such an approach is feasible. Results comparing existing and recursive sparse regression techniques for determining the interconnection structure of a simple complex dynamical network are presented.
Keywords
graph theory; matrix algebra; regression analysis; complex dynamical networks; interconnection graph; interconnection structure estimation; recursive sparse estimator; recursive sparse regression technique; regression techniques; Communities; Estimation; Europe; Least squares approximations; Linear regression; Network topology; Vectors;
fLanguage
English
Publisher
ieee
Conference_Titel
Control Conference (ECC), 2009 European
Conference_Location
Budapest
Print_ISBN
978-3-9524173-9-3
Type
conf
Filename
7074857
Link To Document