Training neural network - Forward Pass

1 view (last 30 days)
Gautami Golani
Gautami Golani on 1 Apr 2019
Answered: Greg Heath on 3 Apr 2019
I have written a code for forward pass of neural network with the following inputs and targets as shown in the code below. A bias is also present and there is one hidden layer (with 2 neurons) and one output layer.
I am not sure whether the code I have written is correct. Could someone tell me if my logic is correct?
X = [0 0; 0 1];
D = [0;1];
X = [ones(1,2); X];
w1 = [0.15, 0.1, 0.3];
w2 = [-0.05, -0.2,-0.25];
w3 = [0.1,0.2, -0.3];
lr = 1;
H = 2;
No = 1;
max_iter = 10;
for iter = 1:max_iter
for j=1:H %Forward Pass
inpH1 = (w1*X);
outH1 = logsig(inpH1);
inpH2 = (w2*X);
outH2 = logsig(inpH2);
output = [ones(1,2); outH1; outH2];
end
for k=1:No
inpO = w3*output;
out_layer = logsig(inpO);
end
Error = 0.5*(D-out_layer').^2
end

Answers (1)

Greg Heath
Greg Heath on 3 Apr 2019
You are very mixed up. I suggest going to the library and finding a good elementary NN book.
Obviously, the appropriate info is available online. However, you will have to search and/or get help from elsewhere.
Sorry,
Greg

Categories

Find more on Deep Learning Toolbox in Help Center and File Exchange

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!