Travel and activity temporal rhythm over a week: results from the BMW survey in Belgium

In this study, two similarity measures are applied to investigate intrapersonal and interpersonal variability of daily activity chains and time use distributions. The first measure for activity chain similarity is based on one-dimensional SAM. The second for individual’s time-use distribution is based on Kullback-Leibler distance. This measure complete the similarity of activity chain in its temporal aspect. Based on the similarity measures of activity patterns, a clustering analysis is conducted to elucidate the relationship between activity patterns and individual’s and household characteristics. The canonical discriminant analysis and canonical correlation analysis are applied to investigate the influence of explanatory variables on cluster membership and the correlations of activity chain types and individual’s sociodemographic variables.