LANDMark: an ensemble approach to the supervised selection of biomarkers in high-throughput sequencing data

Rudar, Josip; Porter, Teresita M.; Wright, Michael; Golding, G. Brian; Hajibabaei, Mehrdad

doi:10.1186/s12859-022-04631-z

BMC Bioinformatics

Table 5 Comparison of classifier generalization performance using the BE dataset before and after recursive feature elimination

From: LANDMark: an ensemble approach to the supervised selection of biomarkers in high-throughput sequencing data

	LANDMark (oracle)	LANDMark (No oracle)	Extra trees	LinearSVC	Logistic regression	Random forest	Ridge regression	SGD (MH)	SGD (SH)
LANDMark (oracle)	0.971 ± 0.002 0.963 ± 0.004	0.001 ± 0.001	− 0.005 ± 0.004	0.002 ± 0.004	− 0.013 ± 0.004	− 0.008 ± 0.004	− 0.006 ± 0.002	− 0.026 ± 0.006*	− 0.018 ± 0.006
LANDMark (no oracle)	0.003 ± 0.002	0.971 ± 0.004 0.960 ± 0.004	− 0.006 ± 0.004	− 0.003 ± 0.004	− 0.013 ± 0.004	− 0.009 ± 0.004	− 0.007 ± 0.002	− 0.027 ± 0.006*	− 0.023 ± 0.005*
Extra trees	0.009 ± 0.004	0.006 ± 0.004	0.966 ± 0.004 0.954 ± 0.005	0.003 ± 0.004	− 0.008 ± 0.004	− 0.004 ± 0.003	− 0.002 ± 0.004	− 0.022 ± 0.005*	− 0.023 ± 0.005*
LinearSVC	0.016 ± 0.003**	− 0.013 ± 0.004**	0.007 ± 0.005	0.968 ± 0.003 0.947 ± 0.004	− 0.0118 ± 0.004	0.006 ± 0.005	− 0.004 ± 0.004	− 0.024 ± 0.006*	− 0.021 ± 0.005*
Logistic regression	0.017 ± 0.005**	0.014 ± 0.004**	0.008 ± 0.005	0.001 ± 0.005	0.958 ± 0.004 0.946 ± 0.005	0.004 ± 0.004	0.006 ± 0.003	− 0.014 ± 0.006	− 0.010 ± 0.006
Random forest	0.003 ± 0.004	0.000 ± 0.003	− 0.006 ± 0.003	− 0.013 ± 0.004	− 0.015 ± 0.004	0.962 ± 0.004 0.960 ± 0.004	0.002 ± 0.004	− 0.018 ± 0.006	− 0.014 + 0.006
Ridge regression	− 0.003 ± 0.004	− 0.006 ± 0.004	− 0.012 ± 0.005	− 0.019 ± 0.004*	− 0.020 ± 0.006*	− 0.006 ± 0.005	0.964 ± 0.003 0.966 ± 0.003	− 0.020 ± 0.006	− 0.016 ± 0.006
SGD (MH)	0.030 ± 0.005**	0.027 ± 0.005	0.021 ± 0.006**	0.014 ± 0.006	0.012 ± 0.006	0.027 ± 0.006**	0.033 ± 0.006**	0.944 ± 0.005 0.933 ± 0.005	0.004 ± 0.007
SGD (SH)	0.062 ± 0.013**	0.059 ± 0.013	0.053 ± 0.014**	0.046 ± 0.014	0.044 ± 0.013	0.059 ± 0.013**	0.064 ± 0.012**	0.032 ± 0.006**	0.948 ± 0.005 0.901 ± 0.012

Models were trained using data from the BE amplicon and generalization performance was measured using the balanced accuracy score. The mean performance of each model before and after recursive feature elimination can be found along the main diagonal. The upper triangle reflects the difference of means between each comparison before recursive feature elimination while the bottom triangle reflects differences in means after recursive feature elimination. A single asterisk is used to represent a statistically significant difference (p ≤ 0.05) in generalization performance which favors classifiers along the rows while statistically significant differences favoring the classifiers along the columns are represented using a double asterisk. The mean and standard error are reported

Back to article page

ISSN: 1471-2105

Contact us

General enquiries: journalsubmissions@springernature.com