| Abstract: | This paper presents a decision treeapproach for predicting smokers' quit intentions usingthe data from the International Tobacco Control FourCountry Survey. Three rule-based classification modelsare generated from three data sets using attributes inrelation to demographics, warning labels, and smokers'beliefs. Both demographic attributes and warning labelattributes are important in predicting smokers' quitintentions. The model's ability to predict smokers' quitintentions is enhanced, if the attributes regardingsmokers' internal motivation and beliefs about quittingare included. |
