Please enable JavaScript.
Coggle requires JavaScript to display documents.
Deep Learning - Coggle Diagram
Deep Learning
Optimizers
2.SGD
3.Mini batch SGD
1.Gradient Decent
4.SGD with Momentum
5.NAG (Nesterov Accelerated Gradient)
5.Adagrad
6.Ada delta &RMS Props
7.ADAM
Activation Function
3.ReLU
4.Leaky ReLU
2.TanH
5.Noise ReLU
1.Sigmoid Function
6.SoftMax
Weight Intialization
3.Intilized the Weight is gaussian/normal initilized
4.Uniform initlization
2.Intilized weight is -ve number
5.Xavier/Glorot
1.Intilized same value for each weight
6.HE intilization
Component
Weight
Activation Function
Input
Output
Training a NN
Intialization of Weight
3.finding all the derivatives with respect to weight
1.Define Loss Function
4.update the weight as per the loss (back propagation)
Loss Function