Comparison of Linear and Nonlinear Regressions for Analysis of Volume Growth of Cryptomeria Trees

Bibliographic Information

Other Title
  • 直線および曲線回帰式によるスギ材積生長量の解析の比較
  • 直線および曲線回帰式によるスギ材積生長量の解析の比較〔英文〕
  • チョクセン オヨビ キョクセン カイキシキ ニ ヨル スギ ザイセキ セイチョ

Search this article

Abstract

In order to select one function among other alternatives to represent growth over time, we proposed the following 3-P criteria: (1) Perfection of curve fitting, (2) Predictability of future growth, and (3) Persistency of regression coefficients over time. Using the total volume growth of cryptomeria trees in Taiwan, the linear transformed function of ln(V)=A+B/Age+C*ln(1+1/Age) was compared to the nonlinear function of V=B0*{1-exp(B1*Age)**B2. The nonlinear model had better fit and closer predic-tions, but its regression coefficients were less persistent than the linear model.

Journal

Details 詳細情報について

Report a problem

Back to top