Lactation Milk Yield Prediction with Possibilistic Logistic Regression Analysis

The logistic regression is a popular method to model the probability of a categorical outcome given as a dependent variable. However, the possibilistic logistic regression can be preferred instead of classical logistic regression when the dependent variable has uncertainity. The aim of this study is to use the possibilistic logistic regression on animal husbandry examining the theoretical foundations of the method based on fuzzy logic approach. A total of 90 cows were enrolled in the study and the average milk yield in 305 days was predicted by animal's weight, breed of the animal, age in lactation, num ber of milkings per day and the milking seasons of cows belonging to different breeds. The Mean Degree of Memberships (MDM) and the Mean of Squared Error (MSE) indices were calculated to decide the goodness of fit of the model.The index values were found as MDM=0.896 and MSE=4.871, respectively. It was shown that the model is fit and is succesfull to predict the average milk yield. It can be concluded that the model can provide the businesses on lactation milk yield production an efficient and accurate prediction results.