Title :
Specification of mixed bilinear time series models
Author :
Kumar, Kuldeep ; Ping Zhang
Author_Institution :
Dept. of Econ. & Stat., Bond Univ., Gold Coast, QLD, Australia
Abstract :
Bilinear models are widely used in signal processing, control theory and in many other information sciences. Although, it is easy to specify simple bilinear time series models using Kumar´s third order moment approach, little work has been done on the specification of mixed bilinear time series models. This paper considers a new class of time series combining Autoregressive Moving Average (ARMA) and bilinear models. Further, we propose a method for specification of such mixed bilinear time series models based on Pade´s approximation and third order moments.
Keywords :
autoregressive moving average processes; bilinear systems; method of moments; signal processing; time series; ARMA; Kumar´s third order moment approach; Pade´s approximation; autoregressive moving average processes; control theory; information sciences; mixed bilinear time series model specification; signal processing; Autoregressive processes; Biological system modeling; Correlation; Predictive models; Signal processing; Signal processing algorithms; Time series analysis; ARIMA Models; Bilinear Models; Pade Approximation; Third order moments;
Conference_Titel :
Signal Processing and Integrated Networks (SPIN), 2015 2nd International Conference on
Conference_Location :
Noida
Print_ISBN :
978-1-4799-5990-7
DOI :
10.1109/SPIN.2015.7095269