On a Flexible Generalized Model Averaging Forecasting of Nonlinear Time Series

No abstract extracted

Key words and phrases: Nonlinear/non-Gaussian time series, Generalized marginal forecast model averaging (GMAFMA), Penalized GMAFMA, Exponential family, Semi-parametric smooth- ing 1. Introduction Forecasting is a fundamental but challenging task in many applications, partic- ularly when it comes to large dimension for time series data that exhibit non-linear and/or non-Gaussian dependence features (Terasvirta, Tjøstheim and Granger, 2010)