Author/Authors :
takma, çiğdem ege üniversitesi - ziraat fakültesi - zootekni bölümü, İzmir, turkey , gevrekçi, yakut ege üniversitesi - ziraat fakültesi - zootekni bölümü, İzmir, turkey , karahan, ahmet erhan ege üniversitesi - ziraat fakültesi - zootekni bölümü, İzmir, turkey , atil, hülya ege üniversitesi - ziraat fakültesi - zootekni bölümü, İzmir, turkey , çevik, muzaffer ege üniversitesi - ziraat fakültesi - zootekni bölümü, İzmir, turkey
Abstract :
decision trees are one of the widely used methods in data mining approaches for data classifying and estimating. The regression tree technique is a decision tree algorithm that examines the relationship between independent variables and each other and the dependent variable and summarizes the results with a tree diagram. In this study, effects of cage, hatching period, line, age of sexual maturity (ASM) and body weight of sexual maturity (BWSM) on egg production were examined by Classification and regression tree analysis (CART). 1980 layers were used from a commercial flock in Turkey. Regression tree analyses showed that the most significant variable affecting egg production was age of sexual maturity and it is followed by hatch trait. Average egg production was estimated 105.805±0.29 for layers that has smaller than or equal to 172.5 day age of sexual maturity, whereas this value was determined to be 73.848±2.47 for layers which has larger age of sexual maturity than 172.5 day. The average egg production for layers in 1^st and 2^nd hatching periods were found 107.33±0.28 whereas average egg production for layers in the 3^rd and 4^th hatching periods were found to be 93.83±1.04.
NaturalLanguageKeyword :
Regression tree , CART (classification and regression tree) , egg production , layer , data mining
