MATLAB: Neural Network Toolbox weighting value

Deep Learning Toolboxneural network

How to define weighting value manually in matlab neural network toolbox?

Thanks

Best Answer

Typically, all of the functions are automatically intialized with random intial weights in train or configure. A major exception is newrb which chooses the training vector with the highest error on a linear classifier (H=0).

See some of my designs by searching on

greg configure

Note that you must use configure before train when designing multiple nets in a loop.

Hope this helps.

Thank you for formally accepting my answer

Greg

Related Solutions

MATLAB: Simple algorithm Exponential weighting

Dear Thijs,

This code may help you on your way:

A=1;
for i=1:10
    if i==1
    x(i) = 1;
    end
    if i>1
        x(i) = x(i-1) + A^(i-1)*x(i-1)
    end
end

Good luck! Christiaan

MATLAB: Frequency based weighting in Simulink

One approach can be to fit a model to that table.

For example;

f = [0.1 0.125 0.16 0.2 0.25 0.315 0.4 0.5];
m = [62.4 97.2 158 243 364 527 708 843]*1e3;
sys = frd(m,f,'FrequencyUnit','Hz');
a = ssest(sys,4);

Then you can use "state space block" or "transfer function block" to implement "a" in simulink.

Just to make note of, this would also weight your signal in every other frequency as well. I do not know of a way to weight only the particular frequencies.

Best Answer

Related Solutions

MATLAB: Simple algorithm Exponential weighting

MATLAB: Frequency based weighting in Simulink

Related Question