开源软件名称：

Neural-Networks

开源软件地址：

https://gitee.com/yuanbo-peng/Neural-Networks

开源软件介绍：

Neural Networks

Author: Yuanbo Peng <[email protected]>
Create Date: 18.03.2019
Project Name: Neural Networks
Revision: 1.0

The EBP Training Algorithm for an MLP Encoder

The project is to implement the Error Back-Propagation (EBP) training algorithm for a multi- layer perceptron (MLP) 4-2-4 encoder using MatLab. Intuitively the structure of the encoder is as shown below:

An input layer with 4 units.
A single hidden layer with 2 units.
An output layer with 4 units.

Each unit has a sigmoid activation function. The task of the encoder is to map the following inputs onto outputs:

Input Pattern	Output Pattern
1, 0, 0, 0	1, 0, 0, 0
0, 1, 0, 0	0, 1, 0, 0
0, 0, 1, 0	0, 0, 1, 0
0, 0, 0, 1	0, 0, 0, 1

Activation Functions

Activation functions are used for a neural network to learn and make sense of some data complicated and Non-linear complex functional mappings between the inputs and response variables. There are several commonly used activation functions to fit different data types better, such as Sigmoid, Tanh, and ReLu etc. In this case, the sigmoid function would be applied.