id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
work_hxud3ocdqbgm7meeqxuk5h3hbq	Mohamed Bouguessa	A practical outlier detection approach for mixed-attribute data	2015	40	.pdf	application/pdf	11480	1370	67	model to identify outliers in mixed-attribute data. Figure 1: Mixed-attribute data set with clustered objects and outliers. outlier detection algorithms can be applied to the entire data set. This process results in a reduced data set based on which other outlier scores109 are then used to estimate outlier scores for objects with mixed attribute. we associate to each data point a two-dimensional outlier score vector ~Vi con-212 Figure 3: The estimated outlier scores in the numerical space for the data objects illustration, Fig. 3 shows the estimated outlier scores in the numerical space253 Figure 4: The estimated outlier scores in the categorical space for the data end, we identify the set of data objects that are associated with the outlier score469 we found that the estimated outlier score vectors in each of the ten data sets are560 Figure 9: Estimated density curves of the numerical outlier scores that correspond to three numerical data sets.	./cache/work_hxud3ocdqbgm7meeqxuk5h3hbq.pdf	./txt/work_hxud3ocdqbgm7meeqxuk5h3hbq.txt