Figure 5 Logistic (17) and bipolar sigmoidal (18) activation functions and their deriva­

doi:10.1201/9781482289862-36

ABSTRACT

Note that 9/ from Eq (7) equals yj. Consequently, and in connection wi th Eqs. (1), (2), and (7), for the multilayer perceptron we can write, yj = cp, = OJ, and for the RBF having Gaussian basis functions, yj = cpy = Gj.

Figure 5 Logistic (17) and bipolar sigmoidal (18) activation functions and their deriva­

ABSTRACT

Figure 5 Logistic (17) and bipolar sigmoidal (18) activation functions and their deriva