Deep Neural Network With L – Layers

This article aims to implement a deep neural network with an arbitrary number of hidden layers each containing different numbers of neurons. We will be implementing this neural net using a few helper functions and at last, we will combine these functions to make the L-layer neural network model.
L – layer deep neural network structure (for understanding) 
 

L – layer neural network

The model’s structure is [LINEAR -> tanh](L-1 times) -> LINEAR -> SIGMOID. i.e., it has L-1 layers using the hyperbolic tangent function as activation function followed by the output layer with a sigmoid activation function. 
More about activation functions
Step by step implementation of the neural network: 
 

Initialize the parameters for the L layers
Implement the forward propagation module
Compute the loss at the final layer
Implement the backward propagation module
Finally, update the parameters
Train the model using existing training dataset
Use trained parameters to test model

Naming conventions followed in the article to prevent confusion: 
 

• Each layer in the network is represented by a set of two parameters W matrix (weight matrix) and b matrix (bias matrix). For layer, i these parameters are represented as Wi and bi respectively.
• The linear output of layer, i is represented as Zi, and the output after activation is represented as Ai. The dimensions of Zi and Ai are the same.

Dimensions of the weights and bias matrices. 
The input layer is of the size (x, m) where m is the number of images. 
 

Layer number	Shape of W	Shape of b	Linear Output	Shape of Activation
Layer 1
Layer 2
:
Layer L – 1
Layer L

Code: Importing all the required python libraries.
 

Initialization: 
 

• We will use random initialization for the weight matrices( to avoid identical output from all neurons in the same layer).
• Zero initialization for the biases.
• The number of neurons in each layer is stored in the layer_dims dictionary with keys as layer number.

Code: 
 

Forward propagation module: 
The Forward propagation module will be completed in three steps. We will complete three functions in this order:
 

• linear_forward (to compute linear output Z for any layer)
• linear_activation_forward where activation will be either tanh or Sigmoid.
• L_model_forward [LINEAR -> tanh](L-1 times) -> LINEAR -> SIGMOID (whole model)

The linear forward module (vectorized over all the examples) computes the following equations: 
Zi = Wi * A(i – 1) + bi Ai = activation_func(Zi)
Code: 
 

We will be using this cost function which will measure the cost for the output layer for all training data.
Code: 
 

Backward Propagation Module: 
Similar to the forward propagation module, we will be implementing three functions in this module too. 
 

• linear_backward (to compute linear output Z for any layer)
• linear_activation_backward where activation will be either tanh or Sigmoid.
• L_model_backward [LINEAR -> tanh](L-1 times) -> LINEAR -> SIGMOID (whole model backward propagation)

For layer i, the linear part is: Zi = Wi * A(i – 1) + bi 
Denoting dZi = we can get dWi, dbi and dA(i – 1) as – 



These equations are formulated using differential calculus and keeping the dimensions of matrices appropriate for matrix dot multiplication using np.dot() function.
Code: Python code for Implementation 
 

Here we will be calculating derivative of sigmoid and tanh functions.Understanding derivation of activation functions 
Code: 
 

L-model-backward: 
Recall that when you implemented the L_model_forward function, at each iteration, you stored a cache that contains (X, W, b, and Z). In the backpropagation module, you will use those variables to compute the gradients.
 

Update Parameters: 
Wi = Wi – a*dWi 
bi = bi – a*dbi 
(where a is an appropriate constant known as learning rate) 
 

Code: Training the model
Now it is time to accumulate all the functions written before to form the final L-layered neural network model. The argument X in L_layer_model will be the training dataset and Y being the corresponding labels. 
 

Code: Implementing the predict function to test the image provided. 
 

Provided layers_dims = [12288, 20, 7, 5, 1] when this model is trained with an appropriate amount of training dataset it is up to 80% accurate on test data. 
The parameters are found after training with an appropriate amount of training dataset. 
 

Testing a custom image 
 

Output with learnt parameters: 
 

y = 1, your L-layer model predicts a Cat picture.