Classification of IRIS by radial basis network

1. Introduction to Radial basis networks

The topology diagram of the radial base network is as follows, its network has three layers, the first layer is the input layer, the second layer is the hidden layer, and the third layer is the output layer. The radial basis function (Gaussian function) is used as the basis function, and the input vector space is transformed into the hidden layer space, and the linear division of the original problem is realized. The core of radial base network is the radial basis function of the hidden layer, which calculates the Euclidean distance between the input vector and the center of the base function, rather than the inner product of the input vector and the weight value. The base function generally uses the Gaussian function.

2. Radial Basis Network Learning algorithm

• There are three parameters that need to be trained in a radial base network
• The center of the base function in the hidden layer
• Standard deviation of base function in hidden layer

• The weights between the hidden layer and the output layer

• Algorithm Core steps

  1. Defining cost functions
  2. Modifying the Base Function Center
  3. Modify Standard deviation
  4. Modify Weight value

3. Experimental simulation

4. Reference source

Source

Copyright NOTICE: This article for Bo Master original article, without Bo Master permission not reproduced.

Classification of IRIS by radial basis network