Detection Of Estrus In Bovine Using Machine Learning

Estrus refers to the reproductive stage of cattle during which ovulation occurs. The day of estrus plays an important role in the reproductive life of a cow. Inseminating cow at perfect time of ovulation was a big deal for farmers. Bovines that undergo Silent Estrus will not show any mounting behaviors or physical signs. So detecting estrus in cows undergoing silent estrus is also a challenging task. The main aim of this study is to propose a simple method to detect the estrus in cow by classifying cows into estrus and non-estrus groups, using machine learning algorithm based on correlating the milk parameters such as fat, pH, SNF, specific gravity, density and the age, amount of milking, frequency of milking and breed of the cow. For estrus cows, milk parameters like pH and fat show high significant change whereas SNF and density shows moderate significant change. The selected parameters are fed into machine learning algorithms like Decision Tree Classifiers. The classification based on the selected parameters specifies higher performance for Decision Tree with accuracy 98%. Thus, Decision Tree classifier is defined as an effective classifier and a simple method to detect estrus in cow by utilizing milk parameters.