Neural Network and Regression Based Model for Cows’ Milk Yield Prediction in Different Climatic Gradients

Abstract

The present study was designed to develop the prediction equations for 305 days fat corrected milk yield on the basis of part periods milk yield, milk component and conformation traits of multigenotype cows. Artificial Neural Network model had the best prediction accuracy across varying environments, though Genetic Function Algorithm had the overall best adequacy for fat corrected milk yield predictions (FCM305d=1036.1-98.3RP+22FY+15.92UC-0.07RUH; Adj R²=0.997; RMSE=30.07; BIC=1997.28).