Neural Networks for Applied Sciences and Engineering--Chapter 3

Chapter 3 Neural Networks for Nonlinear Pattern Recognition

the full article：http://note.youdao.com/noteshare?id=c422ed29a00c8057ee9f64345f40ffec

3.1 Overview and Introduction

3.1.1 Multilayer Perceptron

For simple linear problems,hidden neurons are not required.

3.2 Nonlinear Neurons

3.2.1 Neuron Activation Functions

Two differences:

1,The slope is different;

2,The limit to reaches the bounds is different.

3.3 One-Input Multilayer Nonlinear Networks

3.3.1 Processing with a Single Nonlinear Hidden Neuron

The larger the a1,the steeper the slope of the curve at the boundary point.

3.3.2 Examples:Modeling Cyclical Phenomena with Multiple Nonlinear Neurons

3.3.2.1 Example 1 :Approximating a Square Wave

It is important to how to choose the number of neurons!

3.4 Two-Input Multilayer Perceptron Network

3.4.3 Examples:Two-Dimensional Prediction and Classification

3.4.3.2 Example 2: Two-Dimensional Nonlinear Classification

How many output-neurons and hidden-neurons we need?

Three and six.It is important to help we understand the rule for choosing the number of neurons!