Count Data Modeling About Relationship Between Dubai Housing Sales Transactions and Financial Indicators

Abstract

In this study, illustrating and comparing the performances of count data models such as Poisson, negative binomial (NB), Hurdle and zero-inflated models for the determination of factors affected housing sales in Dubai. Model comparisons are made via Akaike's information criterion (AIC), the Vuong test and examining the residuals. Main purpose of this study is building reliable statistical model for relationship between Dubai housing sales and important financial indicators in Dubai. With this study government, investors and customers would have an idea about housing sales transactions in future according to economic indicators.