Estimation of the factors affecting lactation milk yield of Holstein cattle by the adaptive neuro-fuzzy inference system

The aim of this study is to estimate lactation milk yield of Holstein cattle using first calving age, lactation period, and service period with (ANFIS) adaptive neuro-fuzzy inference system. The input variables for the system in the study were first calving age, lactation period, and service period. The output variable from the system was lactation milk yield. Predicted values obtained from the ANFIS were compared with the observed values. Twenty-seven rule functions are used to obtain the best model and 1000 epochs are used to estimate the accuracy of the training and testing error. The relations between the output and input variables are shown with 3D graphics. R2, RMSE, and MAPE evaluation criteria were used to check the accuracy of the system's estimations. As a result, ANFIS estimates of lactation milk yield were quite close to the observed values and a positive correlation (r: 0.848) was found between them. The results showed that ANFIS can be successfully applied to estimate the lactation milk yield.

Journal

TURKISH JOURNAL OF VETERINARY & ANIMAL SCIENCES

Year

2022

Volume

46

Issue

1

Pages

165-169

ISSN

1300-0128

DOI

10.3906/vet-2112-33

WoS UID

WOS:000753528800001