I would like to know how to change the type of radial basis function in newrbe and the number of hidden layer neurons

Hi, the interpretation of newrbe as follow figure, but i can't understand how to change the type of radial basis function in newrbe and the number of hidden layer neurons. if you have answer, please tell me, thank you very much, have a nice day!

Answer 1

[Jaimin]

Hello @Yi Ma

You can use the “newrb” function instead of the “newrbe” function to modify the radial basis function and adjust the number of neurons in the hidden layer.

I have attached a code snippet to help you understand the “newrb” function better.

% Example data
P = [0 1 2 3 4 5]; % Input data
T = [0.5 1.0 1.5 2.0 2.5 3.0]; % Target data

% Parameters for newrb
goal = 0.0;           % Mean squared error goal
spread = 1.0;         % Spread of the radial basis functions
max_neurons = 10;     % Maximum number of neurons
display_freq = 1;     % Display frequency (set to 0 to turn off)

% Create the RBF network
net = newrb(P, T, goal, spread, max_neurons, display_freq);

% View the network
view(net);

You can refer to the following MathWorks documentation to learn more about the “newrb” function.

https://www.mathworks.com/help/deeplearning/ref/newrb.html

I hope this will be helpful.

[Yi Ma]

thanks your answer, but if i still want to use "newrbe", whether there are some methods to konw the type of radial basis function and the number of hidden layer neurons. best wishes!

[Sam Chak]

Hi @Jaimin,

I believe that @Yi Ma wants to use a custom radial basis function (RBF). The Deep Learning Toolbox uses the default Gaussian RBF. In addition to the widely used Gaussian RBF, the inverse quadratic and inverse multiquadric functions are also known to produce good results. @Yi Ma may want to test a novel RBF. Could you advise on how to achieve that?