Gx wj1 Pj Eii gx w1 Pi

Thus, a competition between the hidden layer neurons is introduced that enables a probabilistic interpretation of classification results. The radial symmetry of the activation function àj (x) in (1) is obviously lost by the normalization in (3). In the following, we refer to this issue by using the term generàlized radial basis functions (GRBF). In [10] such a system is called "Hyper-BF Network."

In a final step, a linear signal propagation of the hidden layer activation is performed to the m neurons of an output làyer by weighted summation,

0 0

Post a comment