MATLAB Answers


Formula for two layer FFNN

Asked by Matthew Clark on 22 Mar 2019
Latest activity Edited by Matthew Clark on 23 Mar 2019
y=b2+LW2_1*tansig(b1+IW1_1*x1); %for one hidden layer
what will be my formula look like if I use 2hidden layers 20 15 neurons?


Sign in to comment.

1 Answer

Answer by Greg Heath
on 23 Mar 2019
 Accepted Answer

y1 = b1 + IW1 * x
y2 = b2 + LW2 * tanh( y1 )
y3 = b3 + LW3 * tanh( y2 )
= b3 + LW3 * tanh( b2 + LW2 * tanh( b1 + IW1 * x ) )
Hope this helps
Thank you for formally accepting my answer

  1 Comment

Thanks prof. Heath,
I am using FFBP NN for my electricity load prediction problem, I have 21 inputs which are normalized to fit the range -0.9 and 0.9,using 2 hidden layers with 20 and 15 nodes, using tansing transfer function in both layers and trainlm algorithm with 0.01 learning rate. My target data is two year historical load. I am getting quite satisfying results on new data. Should I something improve ? I know maybe it is quite weird question from me but I am pretty new to NN.

Sign in to comment.