Fig. 1

Graphical representation of synthetic data: first row, parameter configuration is \(n_p=500\), \(r=0\), \(\rho =1\) and \(a=0.01\) (a), \(a=0.1\) (b), and \(a=1\) (c); second row \(n_p=500\), \(a=0.1\), \(\rho =1\) and \(r=0\) (d), \(r=0.1\) (e), and \(r=0.25\) (f); third row, \(n_p=100\), \(a=0.1\), \(r=0\), \(\rho =1\) (g), \(\rho =3\) (h), and \(\rho =5\) (i). “pos” and “neg” entries in the legend stand for positive and negative class, respectively