id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
work_ahtiosddlbhgtj7wlpis3mjplm	C LEE	An approach to mining the multi-relational imbalanced database	2008	12	.pdf	application/pdf	9031	1297	74	decision tree algorithm to solve the imbalanced problem in a multi-relational database. values and a target class, but a tuple in the non-target relations contains only the attribute values. (Anto, Susumu, & Akira, 2000; Ezawa, Singh, & Norton, 1996; Fawcett & Provost, 1996; Kubat, Holte, & Matwin, 1998; Lawrence, Burns, Back, Tsoi, & Giles, 1998). Although there are many multi-relational database mining classifiers as described in Section 2.1, they can not solve GTIP maintains the original data distribution in each non-target relation by restoring the number of target classes of each tuple to a single one as Thus, in order to design a classifier suitable for the multi-relational database, Mr.G-Tree In this section, we apply two real multi-relational databases to compare the accuracy of the positive and the number of multi-relational data mining classifiers proposed, such as TILDE, FOIL, CrossMine and so on, those Multi-relational g-mean decision tree algorithm Multi-relational g-mean decision tree algorithm	./cache/work_ahtiosddlbhgtj7wlpis3mjplm.pdf	./txt/work_ahtiosddlbhgtj7wlpis3mjplm.txt