Gaussian Mixture Models for Identifying Ultra-Processed Foods

Strickland, Joshua, Dare, Murry and Wang, Wenjia (2026) Gaussian Mixture Models for Identifying Ultra-Processed Foods. In: 21st International Conference on Hybrid Artificial Intelligence Systems, 2026-06-18 - 2026-06-19.

Preview

PDF (Gaussian Mixture Models for Identifying Ultra-Processed Foods) - Published Version Available under License Creative Commons Attribution. Download (346kB) | Preview

Abstract

Ultra-processed foods (UPFs) are industrial formulations of multiple ingredients to make a highly palatable food product and account for nearly half of the average British diet and most developed countries. UPFs are strongly associated with obesity, chronic disease, and increased mortality. The identification of UPFs remains heavily reliant upon expert judgement, making consistent and large-scale classification difficult and resource intensive. Considering that the annotated datasets are limited and inconsistent, This research explores unsupervised machine learning approaches - Gaussian Mixture Models(GMM) to address the challenge. Performance is assessed using metrics including cluster coherence and adjusted rand index. These results of GMM, compared with a baseline method - K-means, demonstrate that the unsupervised learning offers a viable, scalable option for automated food categorisation where labelled data is scarce, potentially enabling real-time dietary analysis tool.

Item Type:	Conference or Workshop Item (Paper)
Uncontrolled Keywords:	machine learning,ultra-processed food,gaussian mixture models,unsupervised learning,nova ultra-processed food classification,sdg 3 - good health and well-being ,/dk/atira/pure/sustainabledevelopmentgoals/good_health_and_well_being
Faculty \ School:	Faculty of Science > School of Computing Sciences
UEA Research Groups:	Faculty of Science > Research Groups > Data Science and AI Faculty of Science > Research Groups > Health Computing
Depositing User:	LivePure Connector
Date Deposited:	16 Jul 2026 09:12
Last Modified:	16 Jul 2026 09:12
URI:	https://ueaeprints.uea.ac.uk/id/eprint/103872
DOI:	10.1007/978-3-032-29292-6_9

Downloads

Actions (login required)

View Item