neural network trained using partical swarm optimization

Question

1 个投票

Hello

in order to resolve some constraints optimization problem, i use neural network trained by pso algorithm.

to this end i try to simulate the matlab code proposed in:

https://www.researchgate.net/publication/305325563_Codes_in_MATLAB_for_Training_Artificial_Neural_Network_using_Particle_Swarm_Optimization

however, it generate the following error:

 Error using network/subsasgn>network_subsasgn (line 551)
net.IW{1,1} must be a 10-by-3 matrix.
 Error in network/subsasgn (line 11)
net = network_subsasgn(net,subscripts,v,netname);
 Error in myfunc (line 18)
net.iw{1,1}=xi; % net.inputWeights{1,1}
 Error in rnn_pso>@(x)myfunc(x,n,m,o,net,inputs,targets)
 Error in rnn_pso (line 37)
        f0(i,1)=fun(x0(i,:));

the size of used data: 150x7 input, 150x3 target.

how can i overcome it?

Did you already test this code? othwhise, i can post the code for do it.

4 个评论
显示 2更早的评论隐藏 2更早的评论

BERGHOUT Tarek 2019-2-3

i used the same code but it works perfectly for any dataset that i used; just check this parameters number of neurons in the hidden layer; nulmber of inputs ; number of outputs ;