Skip to main content
Figure 5 | BMC Bioinformatics

Figure 5

From: Principal component analysis-based unsupervised feature extraction applied to in silico drug discovery for posttraumatic stress disorder-mediated heart disease

Figure 5

Matthews correlation coefficients and F measure for various FE methods applied to the simulated data set with partial mislabeling, indicated in Table 1. Correlation coefficients between mislabeled and true labeling were (a) 0.92, (b) 0.68, and (c) 0.60, as shown in Table 1. Other notations are the same as in Figure 2.

Back to article page