6.深度学习练习：Initialization

本文节选自吴恩达老师《深度学习专项课程》编程作业，在此表示感谢。

课程链接：https://www.deeplearning.ai/deep-learning-specialization/

1 - Neural Network model

2 - Zero initialization

3 - Random initialization（掌握）

4 - He initialization（理解）

To get started, run the following cell to load the packages and the planar dataset you will try to classify.

import numpy as np
import matplotlib.pyplot as plt
import sklearn
import sklearn.datasets
from init_utils import sigmoid, relu, compute_loss, forward_propagation, backward_propagation
from init_utils import update_parameters, predict, load_dataset, plot_decision_boundary, predict_dec%matplotlib inline
plt.rcParams['figure.figsize'] = (7.0, 4.0) # set default size of plots
plt.rcParams['image.interpolation'] = 'nearest'
plt.rcParams['image.cmap'] = 'gray'# load image dataset: blue/red dots in circles
train_X, train_Y, test_X, test_Y = load_dataset()

1 - Neural Network model

You will use a 3-layer neural network (already implemented for you). Here are the initialization methods you will experiment with:

Zeros initialization -- setting initialization = "zeros" in the input argument.
Random initialization -- setting initialization = "random" in the input argument. This initializes the weights to large random values.
He initialization -- setting initialization = "he" in the input argument. This initializes the weights to random values scaled according to a paper by He et al., 2015.

Instructions: Please quickly read over the code below, and run it. In the next part you will implement the three initialization methods that this model() calls.

def model(X, Y, learning_rate = 0.01, num_iterations = 15000, print_cost = True, initialization = "he"):"""Implements a three-layer neural network: LINEAR->RELU->LINEAR->RELU->LINEAR->SIGMOID.Arguments:X -- input data, of shape (2, number of examples)Y -- true "label" vector (containing 0 for red dots; 1 for blue dots), of shape (1, number of examples)learning_rate -- learning rate for gradient descent num_iterations -- number of iterations to run gradient descentprint_cost -- if True, print the cost every 1000 iterationsinitialization -- flag to choose which initialization to use ("zeros","random" or "he")Returns:parameters -- parameters learnt by the model"""grads = {}costs = [] # to keep track of the lossm = X.shape[1] # number of exampleslayers_dims = [X.shape[0], 10, 5, 1]# Initialize parameters dictionary.if initialization == "zeros":parameters = initialize_parameters_zeros(layers_dims)elif initialization == "random":parameters = initialize_parameters_random(layers_dims)elif initialization == "he":parameters = initialize_parameters_he(layers_dims)# Loop (gradient descent)for i in range(0, num_iterations):# Forward propagation: LINEAR -> RELU -> LINEAR -> RELU -> LINEAR -> SIGMOID.a3, cache = forward_propagation(X, parameters)# Losscost = compute_loss(a3, Y)# Backward propagation.grads = backward_propagation(X, Y, cache)# Update parameters.parameters = update_parameters(parameters, grads, learning_rate)# Print the loss every 1000 iterationsif print_cost and i % 1000 == 0:print("Cost after iteration {}: {}".format(i, cost))costs.append(cost)# plot the lossplt.plot(costs)plt.ylabel('cost')plt.xlabel('iterations (per hundreds)')plt.title("Learning rate =" + str(learning_rate))plt.show()return parameters

2 - Zero initialization

There are two types of parameters to initialize in a neural network:

the weight matrices： $(W^{[1]}, W^{[2]}, W^{[3]}, ..., W^{[L-1]}, W^{[L]})$

the bias vectors： $(b^{[1]}, b^{[2]}, b^{[3]}, ..., b^{[L-1]}, b^{[L]})$

Exercise: Implement the following function to initialize all parameters to zeros. You'll see later that this does not work well since it fails to "break symmetry", but lets try it anyway and see what happens. Use np.zeros((..,..)) with the correct shapes.

def initialize_parameters_zeros(layers_dims):"""Arguments:layer_dims -- python array (list) containing the size of each layer.Returns:parameters -- python dictionary containing your parameters "W1", "b1", ..., "WL", "bL":W1 -- weight matrix of shape (layers_dims[1], layers_dims[0])b1 -- bias vector of shape (layers_dims[1], 1)...WL -- weight matrix of shape (layers_dims[L], layers_dims[L-1])bL -- bias vector of shape (layers_dims[L], 1)"""parameters = {}L = len(layers_dims)            # number of layers in the networkfor l in range(1, L):parameters['W' + str(l)] = np.zeros((layers_dims[l], layers_dims[l-1]))parameters['b' + str(l)] = np.zeros((layers_dims[l], 1))return parameters

parameters = initialize_parameters_zeros([3,2,1])
print("W1 = " + str(parameters["W1"]))
print("b1 = " + str(parameters["b1"]))
print("W2 = " + str(parameters["W2"]))
print("b2 = " + str(parameters["b2"]))W1 = [[0. 0. 0.][0. 0. 0.]]
b1 = [[0.][0.]]
W2 = [[0. 0.]]
b2 = [[0.]]parameters = model(train_X, train_Y, initialization = "zeros")
print ("On the train set:")
predictions_train = predict(train_X, train_Y, parameters)
print ("On the test set:")
predictions_test = predict(test_X, test_Y, parameters)

plt.title("Model with Zeros initialization")
axes = plt.gca()
axes.set_xlim([-1.5,1.5])
axes.set_ylim([-1.5,1.5])
plot_decision_boundary(lambda x: predict_dec(parameters, x.T), train_X, np.squeeze(train_Y))

The model is predicting 0 for every example.

In general, initializing all the weights to zero results in the network failing to break symmetry. This means that every neuron in each layer will learn the same thing, and you might as well be training a neural network with ?[?]=1 for every layer, and the network is no more powerful than a linear classifier such as logistic regression.

**What you should remember**: - The weights $W^{[l]}$ should be initialized randomly to break symmetry. - It is however okay to initialize the biases $b^{[l]}$ to zeros. Symmetry is still broken so long as $W^{[l]}$ is initialized randomly.

3 - Random initialization（掌握）

To break symmetry, lets intialize the weights randomly. Following random initialization, each neuron can then proceed to learn a different function of its inputs. In this exercise, you will see what happens if the weights are intialized randomly, but to very large values.

Exercise: Implement the following function to initialize your weights to large random values (scaled by *10) and your biases to zeros. Use np.random.randn(..,..) * 10 for weights and np.zeros((.., ..)) for biases. We are using a fixed np.random.seed(..) to make sure your "random" weights match ours, so don't worry if running several times your code gives you always the same initial values for the parameters.

def initialize_parameters_random(layers_dims):"""Arguments:layer_dims -- python array (list) containing the size of each layer.Returns:parameters -- python dictionary containing your parameters "W1", "b1", ..., "WL", "bL":W1 -- weight matrix of shape (layers_dims[1], layers_dims[0])b1 -- bias vector of shape (layers_dims[1], 1)...WL -- weight matrix of shape (layers_dims[L], layers_dims[L-1])bL -- bias vector of shape (layers_dims[L], 1)"""np.random.seed(3)               # This seed makes sure your "random" numbers will be the as oursparameters = {}L = len(layers_dims)            # integer representing the number of layersfor l in range(1, L):parameters['W' + str(l)] = np.random.randn(layers_dims[l], layers_dims[l-1]) * 10parameters['b' + str(l)] = np.zeros((layers_dims[l], 1))return parameters

parameters = model(train_X, train_Y, initialization = "random")
print ("On the train set:")
predictions_train = predict(train_X, train_Y, parameters)
print ("On the test set:")
predictions_test = predict(test_X, test_Y, parameters)

**In summary**: - Initializing weights to very large random values does not work well. - Hopefully intializing with small random values does better. The important question is: how small should be these random values be? Lets find out in the next part!

4 - He initialization（理解）

Finally, try "He Initialization"; this is named for the first author of He et al., 2015. (If you have heard of "Xavier initialization", this is similar except Xavier initialization uses a scaling factor for the weights ?[?]W[l] of sqrt(1./layers_dims[l-1]) where He initialization would use sqrt(2./layers_dims[l-1]).)

Exercise: Implement the following function to initialize your parameters with He initialization.

Hint: This function is similar to the previous initialize_parameters_random(...). The only difference is that instead of multiplying np.random.randn(..,..) by 10, you will multiply it by $\sqrt{\frac{2}{\text{dimension of the previous layer}}}$ ,which is what He initialization recommends for layers with a ReLU activation.

# GRADED FUNCTION: initialize_parameters_hedef initialize_parameters_he(layers_dims):"""Arguments:layer_dims -- python array (list) containing the size of each layer.Returns:parameters -- python dictionary containing your parameters "W1", "b1", ..., "WL", "bL":W1 -- weight matrix of shape (layers_dims[1], layers_dims[0])b1 -- bias vector of shape (layers_dims[1], 1)...WL -- weight matrix of shape (layers_dims[L], layers_dims[L-1])bL -- bias vector of shape (layers_dims[L], 1)"""np.random.seed(3)parameters = {}L = len(layers_dims) - 1 # integer representing the number of layersfor l in range(1, L + 1):### START CODE HERE ### (≈ 2 lines of code)parameters['W' + str(l)] = np.random.randn(layers_dims[l], layers_dims[l-1]) * np.sqrt(2 / layers_dims[l-1])parameters['b' + str(l)] = np.zeros((layers_dims[l], 1))### END CODE HERE ###return parameters