Alignment-free supervised classification of metagenomes by recursive SVM

BMC Genomics

Table 4 The LOOCV error rates on the simulation data

# of selected features	All	1000	500	200	100	50	30	20	10	5
3-tuple	0.16	-	-	-	-	0.18	0.28	0.28	0.28	NaN
4-tuple	0.14	-	-	0.14	0.18	0.18	0.24	0.34	0.40	NaN
5-tuple	0.06	0.06	0.06	0.08	0.02	0.02	0.02	0.00	0.02	0.12
6-tuple	0.04	0.04	0.04	0.04	0.02	0.00	0.00	0.00	0.00	0.00
7-tuple	0.10	0.00	0.00	0.00	0.00	0.00	0.00	0.00	0.04	0.04
8-tuple	0.26	0.10	0.10	0.14	0.28	0.30	0.40	0.46	0.42	0.46

This table shows the LOOCV error rates when no seed is inserted in group 1 and five seeds of length 6 are inserted in group 2 with the seed density 0.005.
“# of selected” indicates the number of feature selected in each level. “features” indicates the k-tuple feature with the particular value of k. “all” is the situation when all the 4^k features are used. “-” is the feature selection level does not exist for the particular k value. “NaN” means the method failed to converge to a result within a given amount of time.

ISSN: 1471-2164