Studying the Stability of One of Non-Linear Autoregressive Relational Models

Abdulghafoor Salim, Hamed Kalaf
2011 ˜Al-œRafidain journal for computer sciences and mathematics  
Time series are usually built on basic assumptions involving stability, Linearity and normality, these three features are so important in both estimating and building the time series models. The study of time series involves these assumption and how to manipulate the unstable time series on the basis of which the suitable mathematical models fit for these series. In this paper, we suggestion the stationarity of one of the non linear -Autoregressive time series models called rational model has
more » ... ational model has been studied which is a fraction whose numerator is the cosine function and its denominator is an exponential Autoregressive models. The singular point and the limit cycle of the models and its stationarity study have been found by adopting the linear approximation technique.
doi:10.33899/csmj.2011.163625 fatcat:zbk7sw2i7nf63nfye6uazts6cu