Fully connected layers - Python Tutorial

- [Instructor] After a sequence of convolution and pulling operations, Convolutional Neural Networks typically include one or more fully connected layers that perform the final classification or regression. As we discussed in the deep learning with Python foundations course, a fully connected layer is one in which each neuron connects to every neuron in the previous layer, as well as every neuron in the subsequent layer. This dense connectivity allows the network to integrate and interpret the features extracted earlier in the model. Before the outputs from the convolution and pulling layers can be passed to the fully connected layers, they must first be converted into a one dimensional vector. This process is known as "Flattening". For example, flattening the three by three pooled feature map shown on the left results in a one dimensional feature vector with nine elements as shown on the right. The flattened feature vector then serves as the input to the fully connected layers. As we…