How can I give notations to the whole CNN network? Like, if I feed input 'x' to feature extractor, we will get the extracted feature embedding vector at the end that will be passed to the linear classifier and at the end we will apply the softmax? Can any one explain this whole process in the notations.
Asked
Active
Viewed 80 times
0
-
What have you tried so far? – Richard Hardy Mar 22 '22 at 20:26
-
X∈X,X is training set,the x is fed to the feature extractor we will get c= C(x;u)where u is weight tensor then this c will be passed to the linear classifier,g(c;w),w is weight tensor. the predictive probability after softmax function will be p(x)=σ(g(c;w)) – Tariq Hussain Mar 22 '22 at 20:33
-
How would you do it for a different kind of neural network? CNNs are neural networks like any other neural network, as my drawing hopefully demonstrates. – Dave Mar 22 '22 at 21:09
-
Yes, but I wanted professional notation for the CNN – Tariq Hussain Mar 23 '22 at 07:36