Obesity predictive process framework based on dietary patterns from grocery data / Nur’aina Daud

Globally, obese population prevalence is estimated using Body Mass Index (BMI) data from collected surveys and generally analyzed using simple linear regression method. Even the World Health Organization (WHO) uses this simple BMI index to define overweight and obesity. However, this BMI method of e...

Full description

Saved in:

Bibliographic Details
Main Author:	Daud, Nur’aina
Format:	Thesis
Language:	English
Published:	2021
Subjects:	Analysis Online data processing
Online Access:	https://ir.uitm.edu.my/id/eprint/46474/1/46474.pdf
Tags:	Add Tag No Tags, Be the first to tag this record!

id	my-uitm-ir.46474
record_format	uketd_dc
spelling	my-uitm-ir.464742022-02-16T04:43:00Z Obesity predictive process framework based on dietary patterns from grocery data / Nur’aina Daud 2021-04 Daud, Nur’aina Analysis Online data processing Globally, obese population prevalence is estimated using Body Mass Index (BMI) data from collected surveys and generally analyzed using simple linear regression method. Even the World Health Organization (WHO) uses this simple BMI index to define overweight and obesity. However, this BMI method of estimating national obesity population prevalence, although simple, is excessively costly involving huge amount of public health data. The aim of the study is to provide an alternatively cheaper obesity predictive analytics using dietary patterns to overcome the current exorbitantly high operation cost. Exploring the feasibility of using grocery data, these data are assessed from physical receipts collected from the selected household sampling which is generated into nutrition data, and then stored in a food engine (G2NE) developed in this study. Data mining technology is then applied on the tested nutrition data to run the obesity predictive analytics. In this study, BMI data are used as a complementary variable to nutrition variable in estimating the individual nutrition consumption, which are assigned as input data in testing. The prediction modelling is developed by conducting the processes in exploratory phase. The findings from the predictive analytics process in exploratory phase show that the use of 70:30 training and testing data split are the best test options for nutrition dataset based on Area Under the Curve (AUC) performance measurement. From the analysis, obese household shows the patterns that have higher intake percentage of processed food of food pyramid level three compared to non-obese household. The processes in the proposed prediction method involve three different domains which are retail, nutrition and health. In bridging the process involving these three domains, the obesity prediction method in this study is proposed in a form of a process framework (G2NOP Framework). This study has come out with two types of research contributions which are practical contributions from data pre-process phase as well as methodological contributions from the exploratory phase and framework integration phase. The limitations in this study include the nutrition assumptions and the scope of the respondents. Hence, the continuation of the future work is to extend the G2NE with the implementation of grocery and nutrition data sharing from the grocery retailers in which the use of larger scope of the household sampling should be considered. 2021-04 Thesis https://ir.uitm.edu.my/id/eprint/46474/ https://ir.uitm.edu.my/id/eprint/46474/1/46474.pdf text en public phd doctoral Universiti Teknologi MARA Faculty of Computer and Mathematical Sciences Mohd Noor, Nor Laila (Prof. Dr.)
institution	Universiti Teknologi MARA
collection	UiTM Institutional Repository
language	English
advisor	Mohd Noor, Nor Laila (Prof. Dr.)
topic	Analysis Online data processing
spellingShingle	Analysis Online data processing Daud, Nur’aina Obesity predictive process framework based on dietary patterns from grocery data / Nur’aina Daud
description	Globally, obese population prevalence is estimated using Body Mass Index (BMI) data from collected surveys and generally analyzed using simple linear regression method. Even the World Health Organization (WHO) uses this simple BMI index to define overweight and obesity. However, this BMI method of estimating national obesity population prevalence, although simple, is excessively costly involving huge amount of public health data. The aim of the study is to provide an alternatively cheaper obesity predictive analytics using dietary patterns to overcome the current exorbitantly high operation cost. Exploring the feasibility of using grocery data, these data are assessed from physical receipts collected from the selected household sampling which is generated into nutrition data, and then stored in a food engine (G2NE) developed in this study. Data mining technology is then applied on the tested nutrition data to run the obesity predictive analytics. In this study, BMI data are used as a complementary variable to nutrition variable in estimating the individual nutrition consumption, which are assigned as input data in testing. The prediction modelling is developed by conducting the processes in exploratory phase. The findings from the predictive analytics process in exploratory phase show that the use of 70:30 training and testing data split are the best test options for nutrition dataset based on Area Under the Curve (AUC) performance measurement. From the analysis, obese household shows the patterns that have higher intake percentage of processed food of food pyramid level three compared to non-obese household. The processes in the proposed prediction method involve three different domains which are retail, nutrition and health. In bridging the process involving these three domains, the obesity prediction method in this study is proposed in a form of a process framework (G2NOP Framework). This study has come out with two types of research contributions which are practical contributions from data pre-process phase as well as methodological contributions from the exploratory phase and framework integration phase. The limitations in this study include the nutrition assumptions and the scope of the respondents. Hence, the continuation of the future work is to extend the G2NE with the implementation of grocery and nutrition data sharing from the grocery retailers in which the use of larger scope of the household sampling should be considered.
format	Thesis
qualification_name	Doctor of Philosophy (PhD.)
qualification_level	Doctorate
author	Daud, Nur’aina
author_facet	Daud, Nur’aina
author_sort	Daud, Nur’aina
title	Obesity predictive process framework based on dietary patterns from grocery data / Nur’aina Daud
title_short	Obesity predictive process framework based on dietary patterns from grocery data / Nur’aina Daud
title_full	Obesity predictive process framework based on dietary patterns from grocery data / Nur’aina Daud
title_fullStr	Obesity predictive process framework based on dietary patterns from grocery data / Nur’aina Daud
title_full_unstemmed	Obesity predictive process framework based on dietary patterns from grocery data / Nur’aina Daud
title_sort	obesity predictive process framework based on dietary patterns from grocery data / nur’aina daud
granting_institution	Universiti Teknologi MARA
granting_department	Faculty of Computer and Mathematical Sciences
publishDate	2021
url	https://ir.uitm.edu.my/id/eprint/46474/1/46474.pdf
_version_	1783734752024461312

Obesity predictive process framework based on dietary patterns from grocery data / Nur’aina Daud

Similar Items