Modelling Body Mass Index Distribution using Maximum Entropy Density

The objective of this paper is to model the distribution of Body Mass Index (BMI) for a given set of covariates. BMI is one of the leading indicators of health and has been studied by health professionals for many years. As such, there have been various approaches to model the distribution of BMI....

Full description

Bibliographic Details
Main Authors: Singh, Ranjodh, Chan, F., Harris, Mark
Other Authors: Weber, T
Format: Conference Paper
Published: 2015
Online Access:http://hdl.handle.net/20.500.11937/78300
Description
Summary:The objective of this paper is to model the distribution of Body Mass Index (BMI) for a given set of covariates. BMI is one of the leading indicators of health and has been studied by health professionals for many years. As such, there have been various approaches to model the distribution of BMI. Furthermore, there are numerous studies which investigate the association between an individual’s physical and socio-economic attributes (covariates) to their BMI levels. This paper proposes the use of Maximum Entropy Density (MED) to model the distribution of BMI using information from covariates. The paper shows how covariates can be incorporated into the MED framework. This framework is then applied to an Australian data set. The results show how different covariates affect different moments of the estimated BMI distribution.