Application of Vector Autoregressive (VAR) Model and Markov Switching Vector Autoregressive (MSVAR) Model in Modeling Common Cancers Data in Rivers State of Nigeria
Love Cherukei Nnoka *
Department of Statistics, Captain Elechi Amadi Polytechnic, Rumuola, P.M.B-5936, Port Harcourt, Nigeria.
Isaac Didi Essi
Department of Mathematics, Rivers State University, Nkpolu, PMB-5080, Port Harcourt, Nigeria.
Anthony Ike Wegbom
Department of Public Health Sciences, Rivers State University, Nkpolu, P.M.B-5080 Port Harcourt, Nigeria.
*Author to whom correspondence should be addressed.
Abstract
The study investigated the application of VAR and Markov Switching VAR in Modeling Common Cancers Data in Rivers State of Nigeria. The objectives of the study may include the followings: to ascertain the trend in the various types of cancer disease in Rivers state, model and estimate the linear interdependence between the various types of cancer in Rivers State using the Vector Autoregressive (VAR) Model, determine the direction of causality and the significance of causality among the various types of cancer in Rivers state, fit the various types of cancer data to Markov switching Vector Autoregressive (MS-VAR) Model, determine the probability of transition and duration of stay, from one state to another and lastly to compare the estimated Vector Autoregressive (VAR) and Markov switching Vector Autoregressive (MS-VAR) Model based on the model with least Information Criteria. The data for the study was obtained from cancer Registry Department of the University of Port Harcourt Teaching Hospital spanning from January, 2009 to December, 2020, while the software used for data analysis was E-view version twelve. The study adopted an ex-post-facto research design and two major methods were employed in the analysis of data; the Vector Autoregressive Model (VAR) which is a linear modeling technique and the Markov Switching Vector Autoregressive Model (MSVAR) a non-linear modeling technique. The three time series variables were tested for long-run relationship using Johansen co-integration test (the Trace and Max-Eigen Statistics) and the work also investigated stability condition of the fitted model.
Keywords: Vector autoregressive (VAR), cancers, markov