Characterising and predicting car ownership using rough sets

Abstract

This paper applies the relatively new knowledge discovery technique of Rough set analysis to identify the factors that influence the level of car ownership in a household. The study uses the detailed Great Britain National Travel Survey data set which contains information on both household and individual travel behaviour. The knowledge extraction is done using the theory of Rough sets and is presented in the form of easily understood if-then statements or rules which reveal how each attribute influences car ownership behaviour. These rules can then be used to predict household car ownership from information held about previously unseen households and the classification performance of the rules can be assessed. The performance of this classification task is shown to be on a par with other reported studies in this area.