In economic and financial applications, there is often the need for analysing multivariate time series, comprising of time series for a range of quantities. Particularly, in some applications such complex systems can be associated with some underlying network describing pairwise relationships among the quantities. Accounting for the underlying network structure for the analysis of this type of multivariate time series can be particularly informative for forecasting.
This talk will discuss our work, which is motivated by a dataset consisting of time series of firm-to-firm transactions. In this example, pairwise relationships between Standard Industrial Classification (SIC) codes can be represented using a network, while the observed time series for each pair of SIC codes can be regarded as time-varying weights on the edges. Inspired by Knight et al. (2019), we introduce the GNAR-edge model which allows modelling of multiple time series utilising the network structure, assuming that each edge weight depends not only on its past values, but also on past values of its neighbouring edges, for a range of neighbourhood stages. Results from the implementation of the GNAR-edge model on the real firm-to-firm data show good fitting and good predictive performance of the model. In addition, the method is validated through simulations.
Read the related paper at doi.org/10.48550/arXiv.1912.04758
About the speaker:
Dr Anastasia Mantziou is a Postdoctoral Research Associate at The Alan Turing Institute. Prior to that, she was a Research Assistant in statistical cyber-security at Imperial College London. She completed her PhD in Statistics at Lancaster University under the supervision of Dr Simon Lunagomez, Dr Robin Mitra and Professor Paul Fearnhead. Her research interests include network analysis, Bayesian methods and topic modelling. Her research has been applied to networks emerging from various scientific fields such as neuroscience, ecology and computer science (human tracking systems).