Statistical methods to identify mixtures in HBM data



Last Updated: 05-06-2019 13:59


Within task 15.1 different statistical techniques to identify mixtures in HBM data have been applied on a simulated dataset. This presentations shows the methods and results of these different techniques. As a first step in mixture identification, data will be explored based on the correlation structure and principal component analysis. Network analysis uses the partial correlation coefficient to create the most suitable network. These methods will be used to assess the impact of determinants (e.g. differences between countries) on mixtures, through principal component regression and differential network analysis.