| dc.description.abstract | The delivery of consumptive credit by either banks or financial institutions to the right debitur is important. This is done to decrease the loss risk that might be encountered by the banks or financial institutions. CHAID (Chi-square Automatic Interaction Detector) was used in this research, as it is a classification method which generates non-binary classification tree to determine significant factors that affected customer credit payment status. The result of this research showed six from ten predictors was having association with credit status. In this case, CHAID generates 14 segments and potential segments that pointed out qualified customer with well customer payment status, the customer having the characteristics of owning credit card from other banks, has been working for more than ten years, and at least having education of undergraduate or collage (second segment). The classification accuracy table concluded thath potential customers predicted precicely by 98.6%, whereas not qualified customers predicted precicely by 9.3%. The total percentage precice prediction is 80.8%. | en |
