Apuntes
Sin título
Contenido del apunte (1)
Página 1
Suppose a neural network contains a dense layer with that connects one hidden layer h_ihi with n_ini units to a following hidden layer h_{i+1}hi+1 with n_{i+1}ni+1 units. How many parameters are needed for this connection?
(n_i + 1) n_{i+1}(ni+1)ni+1
What is the effect of using pooling layers in convolutional neural networks?
It helps to make the model invariant to small translations of the input.
It reduces the spatial dimensions of the layer input.
Información
- Descripción
- Apunte sobre Sin título, creado por Gustavo Juliano Borges el 07/05/2021.
- Creado por
- Gustavo Juliano Borges
- Visitas
- 0
- Valoración