I. Introduction to Neural Networks

I. Introduction to Neural Networks
The notation for a neural net structure for multiple output neurons is shown in Figure 3. The inputs are x1, x2, x3, ... as before. For the output neurons yin1, yin2, yin3, ... represent the pre-threshold values of output neurons 1, 2, 3, ... with final outputs y1, y2, y3, ... after the threshold function is applied. That is, yj = f(yinj) for each output j, where f is the threshold function. The weight on the connection from input neuron xi to output neuron yj is labeled wij, and the pre-threshold of the jth output neuron is the product of each xi and wij, summed over all i. The objective is to match each output yj to a target value tj.

A more general neural network structure is shown in Figure 4. The layer of neurons between the input and output layers is called a hidden layer. The brain’s structure is composed on many such layers. However, the examples to be shown in this paper will utilize the simpler structure without any hidden layers.