In this study, it was aimed to determine the effect of somatic cell count (SCC) on udder measurements and subclinical mastitis in Holstein cows by data mining method. In the study, the udder measurements and the SCC values of milk samples taken monthly from 79 Holstein cows were used. The Bayesian Net, Decision Table and Nearest Neighbors algorithms were used in the classification of the udder measurements, and model validation is determined by the simple validation method. In the study, it has been found that the best classification model was formed according to the Nearest Neighbors algorithm with the accuracy rate of 97.95% [Root Mean Square Error (RMSE):0.07, Mean Absolute Error (MAE):0.01, Root Relative Squared Error- RRSE (%):22.20, Relative Absolute Error -RAE (%): 5.78, Kappa statistic: 0.95]. The effect of udder measurements on subclinical mastitis was found significant for the front teat length (FTL), the distance between rear teats (DBRT), the distance between side teats (DBST), the rear teat height (RTH) (P<0.01) and the rear teat diameter (RTD) (P<0.05).

Key words: Data mining, Holstein cow, SCC, Subclinical mastitis, Udder measurement