Estimation of a Structural Vector Autoregression Model Using Non-Gaussianity

• DOI: 10.5555/1756006.1859907
• Publication Date: 2010-03-01
• Authors: Aapo Hyvärinen; Kun Zhang; Shohei Shimizu; Patrik O. Hoyer
• Keywords: Identifiability; Vector autoregression; Bayesian vector autoregression

Analysis of causal effects between continuous-valued variables typically uses either autoregressive models or structural equation with instantaneous effects. Estimation Gaussian, linear poses serious identifiability problems, which is why it was recently proposed to use non-Gaussian models. Here, we show how combine the model This effectively what called a vector autoregression (SVAR) model, and thus our work contributes long-standing problem estimate SVAR's. We that such identifiable without prior knowledge network structure. propose computationally efficient methods for estimating as well assess significance influences. The successfully applied on financial brain imaging data.