Identification and estimation of a first order bilinear time series

• Main Author: Afariebor, Roland, 1973-
• Other Authors: Memorial University of Newfoundland. Dept. of Mathematics and Statistics
• Format: Thesis
• Language: English
• Published: 2001
• Subjects: Time-series analysis; Nonlinear theories; Newfoundland studies; University of Newfoundland
• Online Access: http://collections.mun.ca/cdm/ref/collection/theses3/id/96100

Thesis (M.A.S.)--Memorial University of Newfoundland, 2001. Mathematics and Statistics Bibliography: leaves 66-68 In this practicum, we study the properties of a special case of the general bilinear model. The general bilinear model was proposed by Granger and Andersen(1978) and Subba Rao(1981) for studying non-linear time series. Simulation studies and real life data sets are used to evaluate the performance of the theoretical results we derived. The properties we study are the mean, covariance structures, third order moments and cumulants. We find a pattern in the third-order cumulant which is useful in identifying the order of the model. This work is an extension of the result of Oyet(2001). The model is used to make forecasts on three real time series data. -- Also considered are the mean and covariance structures of three other versions of the bilinear model.