Pemilihan Model Terbaik Dalam Analisa Regresi studi Kasus Pendugaan Bobot Hidup sapi Persilangan Simental Dengan sapi PO Berdasarkan Ukuran - Ukuran Tubuh.

Abstract

The objective of the research was to find out the best model of regression equation in predicting body weight of F2 by using body measurement parameters of Simental and Ongole Cross cattle. Eighty two bulls kept by farmer in Padang Panjang, West Sumatra were used in this study. Each bull was measured for their body parantetrs. The parameter measured included body wight (BH), body length (PB), thorax circle (LD), shoulder high (TP) and age in month (US). The data were than statistically analysis by regressin analysis, where the dependent variable was BH, while the independent were PB, LD, TP and US. Criteria for choosing the best equation were R2, R2 Adjusted, Mean square Error (MSE), and Cp Mallow's. The best equation with one independence variable is BH : 6,024 LD - 602,996, the best model with two independence variable is BH : 3,991 PB + 3,809 LD - 775,596, the best model with three independence variable is BH: 3,23 PB + 3,508 LD + 1,855 TP - 852,371 and the best model with four independence variable is BH : 3,20 PB + 3,597 LD + 1,836 TP - 0,181 US - 657,598. The best equation for estimation of body weight was found by using three independence variable of BH: 3,23 PB + 3,508 LD + 1,855 TP - 852,371.