Estimation of Multi-Granger Network Causal Models

• Activity Number: 176
• Type: Contributed
• Date/Time: Monday, August 10, 2015 : 10:30 AM to 12:20 PM
• Sponsor: Business and Economic Statistics Section
• Author(s): Andrey Skripnikov* and George Michailidis
• Companies: University of Florida and University of Florida
• Keywords: network ; Granger ; autoregressive ; VAR ; penalty

Abstract

Network Granger causality focuses on estimating Granger causal effects from p-time series and it can be operationalized through Vector Autoregressive Models (VAR). The latter represent a popular class of time series models that has been widely used in applied econometrics and finance and more recently in biomedical applications. In this work, we discuss joint estimation and model selection issues of multiple Granger causal networks. We present a modeling framework for the setting where the same variables are measured on different entities (e.g. same set of economic activity variables for related countries). The framework involves the introduction of appropriate structural penalties on the transition matrices of the respective VAR models that link the underlying network Granger models and sparse covariance matrices for capturing latent idiosyncratic factors. A fast estimation strategy is presented and the model is evaluated on both synthetic and real macroeconomic data.