PREDICTION OF 305 DAYS MILK YIELD FROM EARLY RECORDS IN DAIRY CATTLE USING ON FUZZY INFERENCE SYSTEM
O. Gorgulu
O. Gorgulu
1 Department of Biostatistics & Medical Informatics, Faculty of Medicine, Ahi Evran University, K1r_ehir,Turkey
Published Online First:
August 01, 2018
Publication Date:
August 01, 2018
ABSTRACT
In the present investigation, Adaptive Neuro Fuzzy Inference System (ANFIS) was implemented to predict 305 d milk yield using partial lactation records of Jersey dairy cattle. The input variables for the system in the study were age, lacta- tion number and milk yields for the first three test-days. The output variable from the system was 305 d milk yield. AN- FIS results related to the milk yields were compared with observed values. Three criteria considered in order to control the reliability of system predictions were the ratio of mean, determination coefficient, and root mean square error. In addition to, the accuracies of ANFIS were compared using the absolute difference between the observed and predicted 2 305 d milk yield. R , RMSE, and RoM values are in the acceptable range. As a conclusion, ANFIS predictions at the beginning of the lactation are related closely to the observed 305 d-lactation yield. The results indicated that ANFIS can be successfully applied for 305 d milk yield early prediction. put using fuzzy logic which then offers a base from which
Keywords:
ANFIS, dairy cow, Fuzzy logic, Inference System, 305d lactation yield