id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
work_lydzvx52yjdtdgjrphenz3g4uy	Katarzyna Bijak	Does segmentation always improve model performance in credit scoring?	2012	33	.pdf	application/pdf	9674	1049	68	Does segmentation always improve model performance in credit scoring? scoring model (scorecard) can be developed for the entire customer population, e.g. found that segmentation does not always improve model performance in credit scoring: for none of the analysed real-world datasets, the multi-scorecard models Credit scoring; Segmentation; Logistic regression; CART; CHAID; LOTUS These models and techniques are used to assess the credit risk of bank customers develop such a multi-scorecard model, segmentation has to be applied. approaches, segmentation is followed by the development of a regression model in In the simultaneous methods, both segmentation and regression models Logistic regression is the most commonly used method for developing scoring models. segmenting customers but also for developing scoring models (Thomas et al., 2002; finding is that segmentation does not always improve model performance in credit The Gini coefficient values of models, scorecards and trees The KS statistic values of models, scorecards and trees 	./cache/work_lydzvx52yjdtdgjrphenz3g4uy.pdf	./txt/work_lydzvx52yjdtdgjrphenz3g4uy.txt