Semi-Supervised Predictive Clustering Trees for (Hierarchical) Multi-Label Classification

Table 2

MLC datasets and their characteristics.

Dataset	Domain
Bibtex [30]	Text	7395	1836/0	159	2.402
Birds [31]	Audio	645	2/258	19	1.014
Emotions [32]	Music	594	0/72	6	1.869
Corel5k [33]	Images	5000	499/0	374	3.522
Enron [34]	Text	1702	1001/0	53	3.378
Genbase [35]	Text	662	1186/0	27	1.252
Mediana [36]	Media	7953	21/58	5	1.205
Medical [37]	Text	978	1449/0	45	1.245
Scene [38]	Images	2407	0/294	6	1.074
SIGMEA real [39]	Ecology	817	0/4	2	0.726
Slovenian rivers [40]	Ecology	1060	0/16	14	5.073
Yeast [41]	Biology	2417	0/103	14	4.237

is the number of examples, is the number of descriptive variables (nominal/continuous), is the number of labels, and is the average number of labels per example.