Mathematical modeling and machine learning prediction for prevalence dynamics of clinical mastitis in dairy herds

Nakov, Dimitar and Zlatanovska, Biljana and Kocaleva, Mirjana and Miteva, Marija and Hristov, Slavča and Stankovic, Branislav (2024) Mathematical modeling and machine learning prediction for prevalence dynamics of clinical mastitis in dairy herds. Zbornik radova 26. medunarodni kongres Mediteranske federacije za zdravlje i produkciju preživara - FeMeSPRum - zbornik radova. p. 22.

Text Mathematical modelling and machine learning prediction.pdf Download (4MB)

Abstract

Mastitis remains one of the major diseases in dairy herds, causing profound economic losses to the entire milk production chain. The main aim of the study was an application of mathematical models and machine learning algorithms for the prediction of mastitis transmission in the dairy cow population. Data used for mathematical models and machine learning algorithms were obtained in a cross-sectional longitudinal survey lasting for one year by analyzing data for clinical mastitis occurrence in three dairy herds. For data prediction, simple SIR and SIRS mathematical models without vital dynamics and Weka software were applied. The annual prevalence rate of clinical mastitis for the entire population of cows was 34.13% on the cow level, 30.07% on the lactation level, while lactation incident risk was 45.86%. Most of the cows manifested one (68.24%) or two (18.63%) cases of clinical mastitis during lactation. The SIR model revealed that after a short time, the epidemic will disappear. From the explanation and the graphical presentations, it can be concluded that the stable point DFE attracts the trajectories of the system. The mastitis on the farms is calming down, and with these parameters of the model, an epidemic cannot occur. With the use of the decision table as one of the most used classification rules and cross-validation folds 10 we can best predict mastitis occurrence in dairy farms. Implementation of a good mastitis prevention program in dairy herds by increasing the rates of control parameters will reduce the mastitis pathogens transmission rates leading to a reduction of mastitis incidence.

Item Type:	Article
Subjects:	Agricultural Sciences > Animal and dairy science Agricultural Sciences > Veterinary science
Divisions:	Faculty of Agriculture
Depositing User:	Dimitar Nakov
Date Deposited:	02 Jul 2024 09:17
Last Modified:	02 Jul 2024 09:17
URI:	https://eprints.ugd.edu.mk/id/eprint/34403

Actions (login required)

View Item