Skip to main content

Advertisement

Figure 5 | BMC Bioinformatics

Figure 5

From: Analysis of nanopore detector measurements using Machine-Learning methods, with application to single-molecule kinetic analysis

Figure 5

(a) SVM performance with different feature sets, for different binary classification data sets: 9AT vs 9TA. The Y-axis, "SN + SP", shows the sum of the Sensitivity and the Specificity. The X-axis is the kernel parameter σ. Note: standard ROC curves are not used here, and in what follows, for two reasons: (i) a comparison with common sigma parameterization was being explored, and (ii) SN and SP are not evaluated individually, whereas (SN + SP) is the measure employed for overall fitness evaluation of a given feature set, kernel, and kernel sigma. (b) SVM performance with 9CG vs 9TA. (c) SVM performance with 9GC vs 9TA. Throughout, the SVM shows that the feature set produced using the inverted emissions performs consistently better than the standard implementation of a HMM.

Back to article page