Selecting the Best Regression Model to State Veneer Volume Recovery
##plugins.themes.academic_pro.article.main##
Abstract
This study was carried out for selecting the best regression equation to state veneer recovery at PT.Tunggal Agathis Indah Wood Industries, Sidangoli, Maluku Utara. Veneer volume recovery is the ratio of veneer volume production by rotary to log volume. The veneer volume recovery is affected by log diameter, physical properties of wood and tree form quotient. The research uses 100 pcs of log from four wood species of Anisoptera spp, Palaqium spp, Canarium spp and Octomeles sumatrana, which are prepared to rotary for peeling. Diameter of log,physical properties, tree form quotient and their interactions are stated as independent variables and the veneer volume recovery is stated as dependent variable. The research uses multiple linear regression analysis for 4, 10, 14 and 15 independent variables. The purpose of this research is to investigate the effect of log dameter, physical properties, tree form quotient (4 variables) and their interactions 10, 14 and 15 variables) on veneer volume recovery, and state the best regression equation to predict it. The best multiple linear regression model for effect of log diameter, moisture content, specific gravity, tree form quotient and their interaction on veneer recovery by stepwise analysis was Y = - 4.82 - 0.002868 X5 + 0.6996 X7 + 0.6686 X8 With R2 = 81.21 %. The veneer volume recovery data is 0.044% bigger than by model.