sepidehamiri / personal-research

0 stars 0 forks source link

Summary of CNN #7

Open sepidehamiri opened 6 years ago

sepidehamiri commented 6 years ago

Three main layers:

1. Convolution Layer

2. Pooling Layers

3. Fully Connected Layers

image

Convolution layers and pooling layers are one among and after them there are some fully connected layers.

feed forward:

The first feed forward just Multiplication point between input and parameter of neuron. Calculation of output Calculation ### loss function

backpropagation

Calculation gradient of parameter with use chain rule

repeat