Machine Learning and Bayesian Modelling Approaches for Big Data in Beef Agriculture & Food Production
Dr Arran Hodgkinson, Queen's University Belfast; Prof Ilias Kyriazakis, Queen’s University Belfast; Dr Mingjun Zhong, University of Aberdeen; Agri-Food and Biosciences Institute
​
Interview date
​
Tuesday 19th November 2024
Apply for this studentship - Applications are now closed for this project.
​
See our Application Page.
​
​
Machine learning (ML) is an invaluable modern technique whose insights, leveraged against large-scale databases tracking agricultural practices and outcomes, could radically change farming practices, both to the benefit of livestock and farmers, alike. As such, we are looking for a motivated candidate with a strong background in data analytics, statistics, and machine learning to take on this exciting project.
Fundamentally, this project intends to utilise cutting-edge ML and probability-based statistical strategies to directly address inefficiencies in feeding practices within the bovine livestock supply chain, employing modelling techniques alongside real-world databases, collected from Northern Irish farms, to create new models of bovine nutritional requirements, based upon metrics commonly used among working farmers. These models will seek to increase sustainability and animal welfare, by facilitating an individual- or subpopulation-based approach to feeding and nutritional consideration.
​
From day one, the candidate will be involved in an exciting collaboration with industrial stakeholders, which will utilise large existing databases to increase our understanding of nutritional needs among bovine livestock and to develop innovative strategies for efficient feeding of demographically diverse populations of cattle. To achieve this, the candidate will learn cutting-edge techniques in data-processing, analysis, and ML and will apply these to develop statistical and bottom-up models (for pharmaceutical industry examples see [3] and [4]), capable of predicting nutritional requirements from obtainable livestock metrics. These results will then feed-forward into farming practices and contribute toward increasing sustainability throughout the agri-food supply-chain; increasing animal welfare; and decreasing costs for farmers and customers.
​
The supervisors supporting this project have expertise spanning population modelling; modern statistical techniques; and addressing inefficiencies and deficits in agri-food supply chains, while the student will have ample opportunities for further study. It is also anticipated that there will exist opportunities for travel to conferences, to present findings, and to facilitate scientific collaboration.
​
​