B0285
Title: Nonlinear network autoregression
Authors: Mirko Armillotta - VU Amsterdam (Netherlands) [presenting]
Konstantinos Fokianos - University of Cyprus (Cyprus)
Abstract: General nonlinear models are studied for time series networks of integer and continuous-valued data. The vector of high dimensional responses, measured on the nodes of a known network, is regressed non-linearly on its lagged value and on lagged values of the neighbouring nodes by employing a smooth link function. Stability conditions are studied for such multivariate processes and develop quasi-maximum likelihood inference when the network dimension is increasing. In addition, linearity score tests are studied by treating separately the cases of identifiable and non-identifiable parameters. In the case of identifiability, the test statistic converges to a chi-square distribution. When the parameters are not identifiable, a supremum-type test is studied whose p-values are approximated adequately by employing a feasible bound and bootstrap methodology. Simulations and data examples complement this.
