Comparative Analysis of Machine Learning Techniques for Predicting Dairy Cow Health Conditions

Main Article Content

Devinder Kaur, Amandeep Kaur

Abstract

Cattle are susceptible to several transmissible diseases that could endanger human health due of our reliance on dairy products. It is imperative to safeguard the welfare of dairy cows to stop these illnesses from spreading. This study compares the effectiveness of various machine learning techniques for identifying disease in dairy cows. A specially designed sensor-based Internet of Things (IoT) gadget has been developed to measure the key health indicators of cows. This innovative tool has been used to gather a comprehensive dataset from 150 cows spread over seven districts in Punjab. Principal component analysis, or PCA, is used to reduce the number of dimensions in the dataset. A variety of techniques, such as Support Vector Machines (SVM), Random Forest (RF), Logistic Regression (LR), and K-Nearest Neighbour (KNN), have been employed to train and assess the models. The models' effectiveness is evaluated by performance indicators that include recall, accuracy, precision, and F1 score. With an accuracy of 99.34%, precision of 97.93%, recall of 97.13%, and F1 score of 97.03%, Random Forest exhibited the best overall results. The results underscore the potential of machine learning, specifically Random Forest, to serve as a reliable tool for early disease detection in dairy cows. This might ultimately enhance animal health and protect public health by limiting the supply of dairy products.

Article Details

Section
Articles