Motif-Based Text Mining of Microbial Metagenome Redundancy Profiling Data for Disease Classification

Table 2

Classification results of pneumonia data in 3-class problem.

Method	Error rate		Dimension	Feature
	On training data	On test data
SVM/FMS	0.1895	0.2637	29	Microbes
SVM/PMF	0.062	0.0756	411	Sequences
SVM/Kruskal-Wallis	0.1187	0.5273	272	Sequences
SVM/Information Gain	0.143	0.2124	12	Sequences
SVM/Chi-square statistic	0.1743	0.5909	1280	Sequences
SVM	0.2187	0.3812	4390	Sequences
NNA/FMS	0.2013	0.3406	112	Microbes
NNA/PMF	0.2152	0.2081	786	Sequences
NNA/Kruskal-Wallis	0.2718	0.3363	85	Sequences
NNA/Information Gain	0.2354	0.4141	39	Sequences
NNA/Chi-square statistic	0.2649	0.3107	69	Sequences
NNA	0.442	0.6162	4390	Sequences