Comparison Between Linear and Non-Linear Regression Models in the Prediction of the Height of Gall Oak Trees (Quercus infectoria Olive.) in Duhok Governorate

Abstract

Abstract This study was conducted for two main purposes, the first one was to study the mathematical relationship between naturally growing gall oak trees in three microsites (Chamankey, Swaratoka, and Baroshky) in Dohuk Governorate and the second one was to compare the ability of linear and nonlinear regression models in height estimation. For such purposes, a sample of 152 trees was purposely selected from these microsites in such a way that they cover the whole variation of breast height diameter in these microsites. The STATGRAPHIC Centurion along with other computer facilities including Excel were used in estimating the parameters of 32 regression models for all microsites together. Many criteria for testing the performance ability of the developed regression equations in the prediction of the height of trees were used in order to select the most appropriate regression model for all microsites together. Screening of the generated models was performed in three stages; the first one is the use of mathematical analysis to see if they have biological limitations. The second stage is to see the performance ability of homogeneous models,(models with the same form of the dependent variable) in the prediction of the dependent variable. The last stage dealt with testing the performance ability of heterogeneous models, (models with different forms of the dependent variable) in the prediction of the dependent variable.