ABSTRACT
The output of a neuron is obtained through the saturation function Y = f(u). In the neural network diagram shown in Fig. 27.20c, the neurons are represented by small circles that
are interconnected to produce a hierarchical arrangement that is capable of pattern recognition. This process of pattern recognition involves learning. During each learning cycle, the weight coefficients Wl' w2, ... wn are adjusted which leads to improved quality of pattern recognition.
