Machine Learning with Apache Spark Quick Start Guide

Output layer

The hidden layer outputs are then used as inputs to calculate the final outputs in the output layer. In our case, we only have a single hidden layer, h¹, with outputs . These then become n inputs into the output layer.

The net input into the activation function for the output layer neurons is then these n inputs computed by the hidden layer and multiplied by a weight set vector, W^h, where each weight set vector, W^h, contains n weights (corresponding to the n hidden layer inputs). For the sake of simplicity, let's assume that we only have one output neuron in our output layer. The weight set vector for this neuron is therefore the ...