Feature Selection and Overlapping Clustering-Based Multilabel Classification Model

Table 8

Hamming loss measure for all models on scene dataset. is the number of attributes, and is the amount of clusters.

Classification model			Scene
-means + ML-KNN	2	294
EM + ML-KNN	2	294
MDI + FCM + ML-KNN	2	20%
		30%
		40%
-means + ML-KNN	3	294
EM + ML-KNN	3	294
MDI + FCM + ML-KNN	3	20%
		30%
		40%
-means + ML-KNN	4	294
EM + ML-KNN	4	294
MDI + FCM + ML-KNN	4	20%
		30%
		40%
-means + ML-KNN	5	294
EM + ML-KNN	5	294
MDI + FCM + ML-KNN	5	20%
		30%
		40%
-means + ML-KNN	6	294
EM + ML-KNN	6	294
MDI + FCM + ML-KNN	6	20%
		30%
		40%
-means + ML-KNN	7	294
EM + ML-KNN	7	294
MDI + FCM + ML-KNN	7	20%
		30%
		40%