Modelling of Networks of Point Processes

Thursday, December 5, 2019 - 3:00pm to 4:00pm

Event Calendar Category

Other LIDS Events

Speaker Name

Professor Victor Solo


School of Electrical Engineering, UNSW, Sydney, Australia, MGH-MIT-HMS Martinos Center for Biomedical Imaging

Building and Room Number



Point Processes have a long history of application in communication network design. But in the last two decades or so a number of significant 'big data' system identification applications have emerged. In neural coding, where spike train data is recorded on a milli-second (ms) time-scale from animal brain neurons in order to study how the brain processes information; in high frequency finance where times of trades are recorded on a ms time-scale and modelled to study financial contagion; in genomics where gene regulatory networks are studied by modelling genes as points along a genome; and more recently in streaming data (social media) where e.g. herding phenomena have been studied and and finally in event triggered control where massive numbers of cheap monitoring sensors only emit signals when environmental variables cross a threshold. Not only areĀ  modelling problems for multivariate point processes more challenging than those of multivariate time series, the available methods are far behind. We give a survey including our recent work while also indicating some challenging issues.


Victor Solo holds the Chair in Systems and Control (2000-2003, 2006-) in the School of Electrical Engineering at UNSW, Sydney, Australia. He was previously a Professor of Statistics (1991-2000) at Macquarie University in Sydney, Australia. He has BS degrees in Mathematics, Statistics (hons.I), and Mechanical Engineering (hons.I). His PhD is in Statistics from the ANU.

He has spent about 40% of his academic career in the United States. From 1980-1985 he was Assistant and then Associate Professor of Statistics, at Harvard University. From 1985-1991 he was Associate Professor of Electrical Engineering at Cornell and then at Johns Hopkins. Most recently (2004 - 2006) he was a Professor of Electrical Engineering and Computer Science at the University of Michigan, Ann Arbor where he also had a quarter appointment in the Statistics Department.

For over 25 years he has had an appointment as a Neuroscientist at the MGH-MIT-HMS Martinos Center for Biomedical Imaging including two periods as visiting Professor of Radiology.

He has been an Associate Editor of three IEEE Journals (TAC, PAMI, SPM) and of the Journal of the American Statistical Association and of the journal, Econometric Theory. He is an IEEE Fellow and a Fellow of the American Statistical Association.

His research interests are in: Econometrics (dynamic factor models), Statistics (time series, point processes), Electrical Engineering (adaptive networks, geometric and topological signal processing, geometric stochastic control, network system identification), Neuro-imaging (fMRI and MEG) and Neural Coding. Current Australian Research Council grants are on Network System Identification and Stochastic Remannian Systems.