Abstract:
Mastitis is an acute disease that mostly occurs in milking cows who experience a red,
painful udder with fever.This disease complex is the result of interaction of various factors associated with the host, pathogens and the environment. The current study aims to
address the significant features from the available disease dataset using machine learning
approaches such as Decision Tree, Logistic Regression and Random Forest.Factors are
categorized as external factors in the given dataset. Secondary data is included in this
research project which is collected from the Anti Bacter research group of ASAB.Data
was gathered from the area of district Rawalpindi in order to address the disease vs.
normal cows and to collect the information of each cow by questionnaire survey from
farmers and then labeled the data. The questionnaire performa was designed on the basis of 28 extrinsic factors i.e mastitis history of cows, bedding material, housing system,
no.of attendees ,management system etc. A total 432 lactating cows data are included
in this study.These cows were examined for mastitis by collecting their milk samples.The
Surf Field Mastitis test (SFMT) was then used in order to classify the disease vs normal cases.In this study,Chi-square test is used to determine the association between
the dependent variables i,e mastitis disease and the independent variables given in the
data.Assessment analysis is performed on the predictive models through accuracy, sensitivity, specificity and precision. ROC curve is used for comparative analysis of predictive
machine learning models.This study would help to spread awareness among farmers.
