¹Bursa Uludağ University, Faculty of Agriculture, Department of Animal Science, TR-16059 Bursa - TÜRKİYE
²Bingöl University, Faculty of Agriculture, Department of Animal Science, TR-12000 Bingöl - TÜRKİYE DOI : 10.9775/kvfd.2022.27149 Th is study was carried out to determine the eff ect of silage type, silage consumption, birth type (single or twin) and birth weight on live weight at the end of fattening in Kıvırcık lambs. In the experiment, 40 male Kıvırcık lambs aged 2.5-3 months were used and the animals were fattened for 56 days. During the fattening period, the lambs fed with 5 diff erent types of silage (100% sunfl ower silage, 75% sunfl ower + 25% corn silage, 50% sunfl ower + 50% corn silage, 25% sunfl ower + 75% corn silage, 100% corn silage) pure and mixed in diff erent proportions and concentrate feed. Data on fattening results were analyzed with MARS and Bagging MARS algorithms. Th e main objective of this research is to predict fattening final live weight (FFLW) of lambs using Multivariate Adaptive Regression Splines (MARS) and Bagging MARS algorithms as a nonparametric regression technique. Live weight value was modeled based on factors such as birth type, birth weight, silage type and silage consumption. Correlation coefficient (r), determination coefficient (R2), Adjust R2, Root-meansquare error (RMSE), standard deviation ratio (SD ratio), mean absolute percentage error (MAPE), mean absolute deviation (MAD), and Akaike Information Criteria (AIC) values of MARS algorithm predicting live weight were as follows: 0.9986, 0.997, 0.977, 0.142, 0.052, 0.2389, 0.086 and -88 respectively. Like statistics for Bagging MARS algorithm were 0.754, 0.556, 0.453, 1.8, 0.666, 3.96, 1.47 and 115 respectively. It was observed that MARS and Bagging MARS algorithms have revealed correct results according to goodness of fit statistics. In this study it has been determined that the MARS algorithm gives better results in live weight modeling. Keywords : Kıvırcık lamb, Silage type, Birth weight, Birth type, Data mining