. Author manuscript; available in PMC: 2018 May 1.

Published in final edited form as: Med Image Anal. 2017 Mar 8;38:104–116. doi: 10.1016/j.media.2017.03.002

Table 3.

Classification accuracy on OutexTC10, OutexTC12000, OutexTC12001, Out-exTC36tl84, OutexTC36horizon, OutexTC11[b, c, n, s] and OutexTC23[b, c, n, s] datasets using different combinations of features and classifiers.

Method	Integrated Features		MRELBP
Datasets	KNN	SVM	KNN	SVM
OutexTC10	0.999768	0.999768	^*0.9984	^-
OutexTC12000	1.0	0.99875	^*0.9949	^-
OutexTC12001	0.999583	0.998541	^*0.9977	^-
OutexTC36tl84	0.956897	0.934863	^*0.9255	^-
OutexTC36horizon	0.960449	0.931632	^*0.9155	^-
OutexTC11b	0.998958	1	^*0.9577	^-
OutexTC11c	0.996875	1	^*0.9472	^-
OutexTC11n	0.99375	0.996875	^*0.877	^-
OutexTC11s	0.983333	0.997916	^*0.9984	^-
OutexTC23b	0.809559	0.917647	^*0.8797	^-
OutexTC23c	0.803308	0.892279	^*0.8036	^-
OutexTC23n	0.816911	0.861397	^*0.664	^-
OutexTC23s	0.759926	0.792279	^*0.9436	^-

: quoted number from (Liu et al., 2016);

: the result is unavailable.