April 2017
Intermediate to advanced
532 pages
12h 39m
English
The neural network is a complex adaptive system, which changes its internal structure based on the information flowing through it using weights. Optimizing the weights of a multilayered neural network is called backpropagation. Backpropagation is a bit beyond the scope of this book, and involves an activation function and basic calculus.
The multilayer perceptron classifier is based on a feed-forward artificial neural network. It consists of multiple layers. Each neural layer is completely connected to the next neural layer in the network, and nodes in the input layer denote the input data. All other nodes map inputs to the outputs by performing a linear combination of the inputs with the nodes' weights and ...
Read now
Unlock full access