University of Twente Student Theses

Login

Detection of subclinical ketosis in dairy cows using behaviour sensor data

Monshouwer, R. (2020) Detection of subclinical ketosis in dairy cows using behaviour sensor data.

Full text not available from this repository.

Full Text Status:Access to this publication is restricted
Embargo date:12 November 2022
Abstract:As precision dairy farming technologies become more common, new possibilities arise in the management of health in dairy cows. Ketosis is a transition cow disease, caused by the start of milk production after calving. Even at the subclinical level, it has an impact on milk yield, reproduction, and other cow diseases causing significant costs on a farm level. Currently, detecting subclinical ketosis by sampling blood is the golden standard, but this is infeasible on a large scale. With behaviour data captured with cow-mounted sensors, large-scale detection is possible. Initial studies to produce a detection model from behaviour data with machine learning have been performed, but a focus on all aspects of the machine learning methodology is lacking. By presenting a complete methodology and comparing time windows, normalisation, features and machine learning models, this study provides an exploratory view on using cow behaviour data to detect subclinical ketosis with machine learning. Using BHBA measurements from two on-farm experiments as targets and behaviour data as input, the detection models are compared. Additionally, a regression variant is proposed to produce a better estimation of subclinical ketosis.
Item Type:Essay (Master)
Clients:
Nedap, Groenlo, Netherlands
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:46 veterinary medicine, 48 agricultural science, 54 computer science
Programme:Computer Science MSc (60300)
Link to this item:http://purl.utwente.nl/essays/85154
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page