On The Analysis Of Bilinear Time Series :- Okposin, Ekanem T.

ABSTRACT

This work is on the analysis of bilinear time series. The bilinear I time series models recognize volterra series. This is why the bilinear time series analysis is used as a paradigm in analysis of nonlinear time series. We consider analysis of bilinear time cummulant analysis. We use the autocorrelation function of the time series process and that of its square transformation to distinguish between a linear and bilinear time series. The result may be useful in differentiating between a MA (q) process and a • and other statistics of pure diagonal bilinear time series, are derived and their stationarity conditions examined. We have established the stationarity condition for variance and third-order cummulant of pure diagonal bilinear time series. Patterns in the third-order cummulant of the pure diagonal bilinear time series pure diagonal bilinear time series model using MINITAB