Increase the odds of gradient descent converging to a lower training (and generalization) error
3种方法
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.
Zero initialization
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 $n^{[l]}=1$ for every layer, and the network is no more powerful than a linear classifier such as logistic regression.
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 network
for l in range(1, L):
### START CODE HERE ### (≈ 2 lines of code)
parameters['W' + str(l)] = np.zeros((layers_dims[l], layers_dims[l-1]))
parameters['b' + str(l)] = 0
### END CODE HERE ###
return parameters
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.
Random initialization
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 ours
parameters = {}
L = len(layers_dims) # integer representing the number of layers
for l in range(1, L):
### START CODE HERE ### (≈ 2 lines of code)
parameters['W' + str(l)] = np.random.randn(layers_dims[l], layers_dims[l-1]) * 10
parameters['b' + str(l)] = np.zeros((layers_dims[l],1))
### END CODE HERE ###
return 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!
He initialization
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]).)
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.
def 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 layers
for 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
What you should remember from this notebook:
Different initializations lead to different results
Random initialization is used to break symmetry and make sure different hidden units can learn different things
Don't intialize to values that are too large
He initialization works well for networks with ReLU activations.
Regularization
L2 Regularization
The standard way to avoid overfitting is called L2 regularization. It consists of appropriately modifying your cost function, from: J = -\frac{1}{m} \sum\limits_{i = 1}^{m} \large{(}\small y^{(i)}\log\left(a^{[L](i)}\right) + (1-y^{(i)})\log\left(1- a^{[L](i)}\right) \large{)} \tag{1} To: J_{regularized} = \small \underbrace{-\frac{1}{m} \sum\limits_{i = 1}^{m} \large{(}\small y^{(i)}\log\left(a^{[L](i)}\right) + (1-y^{(i)})\log\left(1- a^{[L](i)}\right) \large{)} }_\text{cross-entropy cost} + \underbrace{\frac{1}{m} \frac{\lambda}{2} \sum\limits_l\sum\limits_k\sum\limits_j W_{k,j}^{[l]2} }_\text{L2 regularization cost} \tag{2}
Let's modify your cost and observe the consequences.
def compute_cost_with_regularization(A3, Y, parameters, lambd):
"""
Implement the cost function with L2 regularization. See formula (2) above.
Arguments:
A3 -- post-activation, output of forward propagation, of shape (output size, number of examples)
Y -- "true" labels vector, of shape (output size, number of examples)
parameters -- python dictionary containing parameters of the model
Returns:
cost - value of the regularized loss function (formula (2))
"""
m = Y.shape[1]
W1 = parameters["W1"]
W2 = parameters["W2"]
W3 = parameters["W3"]
cross_entropy_cost = compute_cost(A3, Y) # This gives you the cross-entropy part of the cost
### START CODE HERE ### (approx. 1 line)
L2_regularization_cost = 1/m * lambd/2 * (np.sum(np.square(W1))+np.sum(np.square(W2))+np.sum(np.square(W3)))
### END CODER HERE ###
cost = cross_entropy_cost + L2_regularization_cost
return cost
Of course, because you changed the cost, you have to change backward propagation as well! All the gradients have to be computed with respect to this new cost.
Implement the changes needed in backward propagation to take into account regularization. The changes only concern dW1, dW2 and dW3. For each, you have to add the regularization term's gradient ($\frac{d}{dW} ( \frac{1}{2}\frac{\lambda}{m} W^2) = \frac{\lambda}{m} W$).
def backward_propagation_with_regularization(X, Y, cache, lambd):
"""
Implements the backward propagation of our baseline model to which we added an L2 regularization.
Arguments:
X -- input dataset, of shape (input size, number of examples)
Y -- "true" labels vector, of shape (output size, number of examples)
cache -- cache output from forward_propagation()
lambd -- regularization hyperparameter, scalar
Returns:
gradients -- A dictionary with the gradients with respect to each parameter, activation and pre-activation variables
"""
m = X.shape[1]
(Z1, A1, W1, b1, Z2, A2, W2, b2, Z3, A3, W3, b3) = cache
dZ3 = A3 - Y
### START CODE HERE ### (approx. 1 line)
dW3 = 1./m * np.dot(dZ3, A2.T) + lambd/m * W3
### END CODE HERE ###
db3 = 1./m * np.sum(dZ3, axis=1, keepdims = True)
dA2 = np.dot(W3.T, dZ3)
dZ2 = np.multiply(dA2, np.int64(A2 > 0))
### START CODE HERE ### (approx. 1 line)
dW2 = 1./m * np.dot(dZ2, A1.T) + lambd/m * W2
### END CODE HERE ###
db2 = 1./m * np.sum(dZ2, axis=1, keepdims = True)
dA1 = np.dot(W2.T, dZ2)
dZ1 = np.multiply(dA1, np.int64(A1 > 0))
### START CODE HERE ### (approx. 1 line)
dW1 = 1./m * np.dot(dZ1, X.T) + lambd/m * W1
### END CODE HERE ###
db1 = 1./m * np.sum(dZ1, axis=1, keepdims = True)
gradients = {"dZ3": dZ3, "dW3": dW3, "db3": db3,"dA2": dA2,
"dZ2": dZ2, "dW2": dW2, "db2": db2, "dA1": dA1,
"dZ1": dZ1, "dW1": dW1, "db1": db1}
return gradients
Observations:
The value of $\lambda$ is a hyperparameter that you can tune using a dev set.
L2 regularization makes your decision boundary smoother. If $\lambda$ is too large, it is also possible to "oversmooth", resulting in a model with high bias.
What is L2-regularization actually doing?:
L2-regularization relies on the assumption that a model with small weights is simpler than a model with large weights. Thus, by penalizing the square values of the weights in the cost function you drive all the weights to smaller values. It becomes too costly for the cost to have large weights! This leads to a smoother model in which the output changes more slowly as the input changes.
What you should remember -- the implications of L2-regularization on:
The cost computation:
A regularization term is added to the cost
The backpropagation function:
There are extra terms in the gradients with respect to weight matrices
Weights end up smaller ("weight decay"):
Weights are pushed to smaller values.
Dropout
Finally, dropout is a widely used regularization technique that is specific to deep learning. It randomly shuts down some neurons in each iteration.
When you shut some neurons down, you actually modify your model. The idea behind drop-out is that at each iteration, you train a different model that uses only a subset of your neurons. With dropout, your neurons thus become less sensitive to the activation of one other specific neuron, because that other neuron might be shut down at any time.
Forward propagation with dropout
Implement the forward propagation with dropout. You are using a 3 layer neural network, and will add dropout to the first and second hidden layers. We will not apply dropout to the input layer or output layer.
Instructions: You would like to shut down some neurons in the first and second layers. To do that, you are going to carry out 4 Steps:
In lecture, we dicussed creating a variable $d^{[1]}$ with the same shape as $a^{[1]}$ using np.random.rand() to randomly get numbers between 0 and 1. Here, you will use a vectorized implementation, so create a random matrix $D^{[1]} = [d^{1} d^{1} ... d^{1}] $ of the same dimension as $A^{[1]}$.
Set each entry of $D^{[1]}$ to be 0 with probability (1-keep_prob) or 1 with probability (keep_prob), by thresholding values in $D^{[1]}$ appropriately. Hint: to set all the entries of a matrix X to 0 (if entry is less than 0.5) or 1 (if entry is more than 0.5) you would do: X = (X < 0.5). Note that 0 and 1 are respectively equivalent to False and True.
Set $A^{[1]}$ to $A^{[1]} * D^{[1]}$. (You are shutting down some neurons). You can think of $D^{[1]}$ as a mask, so that when it is multiplied with another matrix, it shuts down some of the values.
Divide $A^{[1]}$ by keep_prob. By doing this you are assuring that the result of the cost will still have the same expected value as without drop-out. (This technique is also called inverted dropout.)
def forward_propagation_with_dropout(X, parameters, keep_prob = 0.5):
"""
Implements the forward propagation: LINEAR -> RELU + DROPOUT -> LINEAR -> RELU + DROPOUT -> LINEAR -> SIGMOID.
Arguments:
X -- input dataset, of shape (2, number of examples)
parameters -- python dictionary containing your parameters "W1", "b1", "W2", "b2", "W3", "b3":
W1 -- weight matrix of shape (20, 2)
b1 -- bias vector of shape (20, 1)
W2 -- weight matrix of shape (3, 20)
b2 -- bias vector of shape (3, 1)
W3 -- weight matrix of shape (1, 3)
b3 -- bias vector of shape (1, 1)
keep_prob - probability of keeping a neuron active during drop-out, scalar
Returns:
A3 -- last activation value, output of the forward propagation, of shape (1,1)
cache -- tuple, information stored for computing the backward propagation
"""
np.random.seed(1)
# retrieve parameters
W1 = parameters["W1"]
b1 = parameters["b1"]
W2 = parameters["W2"]
b2 = parameters["b2"]
W3 = parameters["W3"]
b3 = parameters["b3"]
# LINEAR -> RELU -> LINEAR -> RELU -> LINEAR -> SIGMOID
Z1 = np.dot(W1, X) + b1
A1 = relu(Z1)
### START CODE HERE ### (approx. 4 lines) # Steps 1-4 below correspond to the Steps 1-4 described above.
D1 = np.random.rand(A1.shape[0], A1.shape[1]) # Step 1: initialize matrix D1 = np.random.rand(..., ...)
D1 = D1 < keep_prob # Step 2: convert entries of D1 to 0 or 1 (using keep_prob as the threshold)
A1 = np.multiply(D1, A1) # Step 3: shut down some neurons of A1
A1 = A1 / keep_prob # Step 4: scale the value of neurons that haven't been shut down
### END CODE HERE ###
Z2 = np.dot(W2, A1) + b2
A2 = relu(Z2)
### START CODE HERE ### (approx. 4 lines)
D2 = np.random.rand(A2.shape[0], A2.shape[1]) # Step 1: initialize matrix D2 = np.random.rand(..., ...)
D2 = D2 < keep_prob # Step 2: convert entries of D2 to 0 or 1 (using keep_prob as the threshold)
A2 = np.multiply(D2, A2) # Step 3: shut down some neurons of A2
A2 = A2 / keep_prob # Step 4: scale the value of neurons that haven't been shut down
### END CODE HERE ###
Z3 = np.dot(W3, A2) + b3
A3 = sigmoid(Z3)
cache = (Z1, D1, A1, W1, b1, Z2, D2, A2, W2, b2, Z3, A3, W3, b3)
return A3, cache
Backward propagation with dropout
Implement the backward propagation with dropout. As before, you are training a 3 layer network. Add dropout to the first and second hidden layers, using the masks $D^{[1]}$ and $D^{[2]}$ stored in the cache.
Instruction: Backpropagation with dropout is actually quite easy. You will have to carry out 2 Steps:
You had previously shut down some neurons during forward propagation, by applying a mask $D^{[1]}$ to A1. In backpropagation, you will have to shut down the same neurons, by reapplying the same mask $D^{[1]}$ to dA1.
During forward propagation, you had divided A1 by keep_prob. In backpropagation, you'll therefore have to divide dA1 by keep_prob again (the calculus interpretation is that if $A^{[1]}$ is scaled by keep_prob, then its derivative $dA^{[1]}$ is also scaled by the same keep_prob).
def backward_propagation_with_dropout(X, Y, cache, keep_prob):
"""
Implements the backward propagation of our baseline model to which we added dropout.
Arguments:
X -- input dataset, of shape (2, number of examples)
Y -- "true" labels vector, of shape (output size, number of examples)
cache -- cache output from forward_propagation_with_dropout()
keep_prob - probability of keeping a neuron active during drop-out, scalar
Returns:
gradients -- A dictionary with the gradients with respect to each parameter, activation and pre-activation variables
"""
m = X.shape[1]
(Z1, D1, A1, W1, b1, Z2, D2, A2, W2, b2, Z3, A3, W3, b3) = cache
dZ3 = A3 - Y
dW3 = 1./m * np.dot(dZ3, A2.T)
db3 = 1./m * np.sum(dZ3, axis=1, keepdims = True)
dA2 = np.dot(W3.T, dZ3)
### START CODE HERE ### (≈ 2 lines of code)
dA2 = np.multiply(dA2, D2) # Step 1: Apply mask D2 to shut down the same neurons as during the forward propagation
dA2 = dA2 / keep_prob # Step 2: Scale the value of neurons that haven't been shut down
### END CODE HERE ###
dZ2 = np.multiply(dA2, np.int64(A2 > 0))
dW2 = 1./m * np.dot(dZ2, A1.T)
db2 = 1./m * np.sum(dZ2, axis=1, keepdims = True)
dA1 = np.dot(W2.T, dZ2)
### START CODE HERE ### (≈ 2 lines of code)
dA1 = np.multiply(dA1, D1) # Step 1: Apply mask D1 to shut down the same neurons as during the forward propagation
dA1 = dA1 / keep_prob # Step 2: Scale the value of neurons that haven't been shut down
### END CODE HERE ###
dZ1 = np.multiply(dA1, np.int64(A1 > 0))
dW1 = 1./m * np.dot(dZ1, X.T)
db1 = 1./m * np.sum(dZ1, axis=1, keepdims = True)
gradients = {"dZ3": dZ3, "dW3": dW3, "db3": db3,"dA2": dA2,
"dZ2": dZ2, "dW2": dW2, "db2": db2, "dA1": dA1,
"dZ1": dZ1, "dW1": dW1, "db1": db1}
return gradients
Note:
A common mistake when using dropout is to use it both in training and testing. You should use dropout (randomly eliminate nodes) only in training.
Deep learning frameworks like tensorflow, PaddlePaddle, keras or caffe come with a dropout layer implementation. Don't stress - you will soon learn some of these frameworks.
What you should remember about dropout:
Dropout is a regularization technique.
You only use dropout during training. Don't use dropout (randomly eliminate nodes) during test time.
Apply dropout both during forward and backward propagation.
During training time, divide each dropout layer by keep_prob to keep the same expected value for the activations. For example, if keep_prob is 0.5, then we will on average shut down half the nodes, so the output will be scaled by 0.5 since only the remaining half are contributing to the solution. Dividing by 0.5 is equivalent to multiplying by 2. Hence, the output now has the same expected value. You can check that this works even when keep_prob is other values than 0.5.
Gradient Checking
1) How does gradient checking work?
Backpropagation computes the gradients $\frac{\partial J}{\partial \theta}$, where $\theta$ denotes the parameters of the model. $J$ is computed using forward propagation and your loss function.
Because forward propagation is relatively easy to implement, you're confident you got that right, and so you're almost 100% sure that you're computing the cost $J$ correctly. Thus, you can use your code for computing $J$ to verify the code for computing $\frac{\partial J}{\partial \theta}$.
Let's look back at the definition of a derivative (or gradient): \frac{\partial J}{\partial \theta} = \lim_{\varepsilon \to 0} \frac{J(\theta + \varepsilon) - J(\theta - \varepsilon)}{2 \varepsilon} \tag{1}
If you're not familiar with the "$\displaystyle \lim_{\varepsilon \to 0}$" notation, it's just a way of saying "when $\varepsilon$ is really really small."
We know the following:
$\frac{\partial J}{\partial \theta}$ is what you want to make sure you're computing correctly.
You can compute $J(\theta + \varepsilon)$ and $J(\theta - \varepsilon)$ (in the case that $\theta$ is a real number), since you're confident your implementation for $J$ is correct.
Lets use equation (1) and a small value for $\varepsilon$ to convince your CEO that your code for computing $\frac{\partial J}{\partial \theta}$ is correct!
2) 1-dimensional gradient checking
Instructions:
First compute "gradapprox" using the formula above (1) and a small value of $\varepsilon$. Here are the Steps to follow:
Then compute the gradient using backward propagation, and store the result in a variable "grad"
Finally, compute the relative difference between "gradapprox" and the "grad" using the following formula: difference = \frac {\mid\mid grad - gradapprox \mid\mid_2}{\mid\mid grad \mid\mid_2 + \mid\mid gradapprox \mid\mid_2} \tag{2} You will need 3 Steps to compute this formula:
1'. compute the numerator using np.linalg.norm(...)
2'. compute the denominator. You will need to call np.linalg.norm(...) twice.
3'. divide them.
If this difference is small (say less than $10^{-7}$), you can be quite confident that you have computed your gradient correctly. Otherwise, there may be a mistake in the gradient computation.
def gradient_check(x, theta, epsilon = 1e-7):
"""
Implement the backward propagation presented in Figure 1.
Arguments:
x -- a real-valued input
theta -- our parameter, a real number as well
epsilon -- tiny shift to the input to compute approximated gradient with formula(1)
Returns:
difference -- difference (2) between the approximated gradient and the backward propagation gradient
"""
# Compute gradapprox using left side of formula (1). epsilon is small enough, you don't need to worry about the limit.
### START CODE HERE ### (approx. 5 lines)
thetaplus = theta + epsilon # Step 1
thetaminus = theta - epsilon # Step 2
J_plus = thetaplus * x # Step 3
J_minus = thetaminus * x # Step 4
gradapprox = (J_plus - J_minus)/(2 * epsilon) # Step 5
### END CODE HERE ###
# Check if gradapprox is close enough to the output of backward_propagation()
### START CODE HERE ### (approx. 1 line)
grad = backward_propagation(x, theta)
### END CODE HERE ###
### START CODE HERE ### (approx. 1 line)
numerator = np.linalg.norm(grad - gradapprox) # Step 1'
denominator = np.linalg.norm(grad) + np.linalg.norm(gradapprox) # Step 2'
difference = numerator / denominator # Step 3'
### END CODE HERE ###
if difference < 1e-7:
print ("The gradient is correct!")
else:
print ("The gradient is wrong!")
return difference
def forward_propagation(x, theta):
"""
Implement the linear forward propagation (compute J) presented in Figure 1 (J(theta) = theta * x)
Arguments:
x -- a real-valued input
theta -- our parameter, a real number as well
Returns:
J -- the value of function J, computed using the formula J(theta) = theta * x
"""
### START CODE HERE ### (approx. 1 line)
J = theta * x
### END CODE HERE ###
return J
def backward_propagation(x, theta):
"""
Computes the derivative of J with respect to theta (see Figure 1).
Arguments:
x -- a real-valued input
theta -- our parameter, a real number as well
Returns:
dtheta -- the gradient of the cost with respect to theta
"""
### START CODE HERE ### (approx. 1 line)
dtheta = x
### END CODE HERE ###
return dtheta
3) N-dimensional gradient checking
How does gradient checking work?.
As in 1) and 2), you want to compare "gradapprox" to the gradient computed by backpropagation. The formula is still:
However, $\theta$ is not a scalar anymore. It is a dictionary called "parameters". We implemented a function "dictionary_to_vector()" for you. It converts the "parameters" dictionary into a vector called "values", obtained by reshaping all parameters (W1, b1, W2, b2, W3, b3) into vectors and concatenating them.
The inverse function is "vector_to_dictionary" which outputs back the "parameters" dictionary.
Figure 2 : dictionary_to_vector() and vector_to_dictionary()
You will need these functions in gradient_check_n()
We have also converted the "gradients" dictionary into a vector "grad" using gradients_to_vector(). You don't need to worry about that.
Instructions: Here is pseudo-code that will help you implement the gradient check.
For each i in num_parameters:
To compute J_plus[i]:
Set $\theta^{+}$ to np.copy(parameters_values)
Set $\theta^{+}_i$ to $\theta^{+}_i + \varepsilon$
Calculate $J^{+}_i$ using to forward_propagation_n(x, y, vector_to_dictionary($\theta^{+}$ )).
To compute J_minus[i]: do the same thing with $\theta^{-}$
Thus, you get a vector gradapprox, where gradapprox[i] is an approximation of the gradient with respect to parameter_values[i]. You can now compare this gradapprox vector to the gradients vector from backpropagation. Just like for the 1D case (Steps 1', 2', 3'), compute: difference = \frac {\| grad - gradapprox \|_2}{\| grad \|_2 + \| gradapprox \|_2 } \tag{3}
def gradient_check_n(parameters, gradients, X, Y, epsilon = 1e-7):
"""
Checks if backward_propagation_n computes correctly the gradient of the cost output by forward_propagation_n
Arguments:
parameters -- python dictionary containing your parameters "W1", "b1", "W2", "b2", "W3", "b3":
grad -- output of backward_propagation_n, contains gradients of the cost with respect to the parameters.
x -- input datapoint, of shape (input size, 1)
y -- true "label"
epsilon -- tiny shift to the input to compute approximated gradient with formula(1)
Returns:
difference -- difference (2) between the approximated gradient and the backward propagation gradient
"""
# Set-up variables
parameters_values, _ = dictionary_to_vector(parameters)
grad = gradients_to_vector(gradients)
num_parameters = parameters_values.shape[0]
J_plus = np.zeros((num_parameters, 1))
J_minus = np.zeros((num_parameters, 1))
gradapprox = np.zeros((num_parameters, 1))
# Compute gradapprox
for i in range(num_parameters):
# Compute J_plus[i]. Inputs: "parameters_values, epsilon". Output = "J_plus[i]".
# "_" is used because the function you have to outputs two parameters but we only care about the first one
### START CODE HERE ### (approx. 3 lines)
thetaplus = np.copy(parameters_values) # Step 1
thetaplus[i][0] = thetaplus[i][0] + epsilon # Step 2
J_plus[i], _ = forward_propagation_n(X, Y, vector_to_dictionary(thetaplus)) # Step 3
### END CODE HERE ###
# Compute J_minus[i]. Inputs: "parameters_values, epsilon". Output = "J_minus[i]".
### START CODE HERE ### (approx. 3 lines)
thetaminus = np.copy(parameters_values) # Step 1
thetaminus[i][0] = thetaminus[i][0] - epsilon # Step 2
J_minus[i], _ = forward_propagation_n(X, Y, vector_to_dictionary(thetaminus)) # Step 3
### END CODE HERE ###
# Compute gradapprox[i]
### START CODE HERE ### (approx. 1 line)
gradapprox[i] = (J_plus[i] - J_minus[i]) / (2 * epsilon)
### END CODE HERE ###
# Compare gradapprox to backward propagation gradients by computing difference.
### START CODE HERE ### (approx. 1 line)
numerator = np.linalg.norm(gradapprox - grad) # Step 1'
denominator = np.linalg.norm(grad) + np.linalg.norm(gradapprox) # Step 2'
difference = numerator / denominator # Step 3'
### END CODE HERE ###
if difference > 1e-7:
print ("\033[93m" + "There is a mistake in the backward propagation! difference = " + str(difference) + "\033[0m")
else:
print ("\033[92m" + "Your backward propagation works perfectly fine! difference = " + str(difference) + "\033[0m")
return difference
def dictionary_to_vector(parameters):
"""
Roll all our parameters dictionary into a single vector satisfying our specific required shape.
"""
keys = []
count = 0
for key in ["W1", "b1", "W2", "b2", "W3", "b3"]:
# flatten parameter
new_vector = np.reshape(parameters[key], (-1,1))
keys = keys + [key]*new_vector.shape[0]
if count == 0:
theta = new_vector
else:
theta = np.concatenate((theta, new_vector), axis=0)
count = count + 1
return theta, keys
def vector_to_dictionary(theta):
"""
Unroll all our parameters dictionary from a single vector satisfying our specific required shape.
"""
parameters = {}
parameters["W1"] = theta[:20].reshape((5,4))
parameters["b1"] = theta[20:25].reshape((5,1))
parameters["W2"] = theta[25:40].reshape((3,5))
parameters["b2"] = theta[40:43].reshape((3,1))
parameters["W3"] = theta[43:46].reshape((1,3))
parameters["b3"] = theta[46:47].reshape((1,1))
return parameters
def gradients_to_vector(gradients):
"""
Roll all our gradients dictionary into a single vector satisfying our specific required shape.
"""
count = 0
for key in ["dW1", "db1", "dW2", "db2", "dW3", "db3"]:
# flatten parameter
new_vector = np.reshape(gradients[key], (-1,1))
if count == 0:
theta = new_vector
else:
theta = np.concatenate((theta, new_vector), axis=0)
count = count + 1
return theta
##Note
Gradient Checking is slow! Approximating the gradient with $\frac{\partial J}{\partial \theta} \approx \frac{J(\theta + \varepsilon) - J(\theta - \varepsilon)}{2 \varepsilon}$ is computationally costly. For this reason, we don't run gradient checking at every iteration during training. Just a few times to check if the gradient is correct.
Gradient Checking, at least as we've presented it, doesn't work with dropout. You would usually run the gradient check algorithm without dropout to make sure your backprop is correct, then add dropout.
What you should remember from this notebook:
Gradient checking verifies closeness between the gradients from backpropagation and the numerical approximation of the gradient (computed using forward propagation).
Gradient checking is slow, so we don't run it in every iteration of training. You would usually run it only to make sure your code is correct, then turn it off and use backprop for the actual learning process.
Optimization Methods
1 - Gradient Descent
def update_parameters_with_gd(parameters, grads, learning_rate):
"""
Update parameters using one step of gradient descent
Arguments:
parameters -- python dictionary containing your parameters to be updated:
parameters['W' + str(l)] = Wl
parameters['b' + str(l)] = bl
grads -- python dictionary containing your gradients to update each parameters:
grads['dW' + str(l)] = dWl
grads['db' + str(l)] = dbl
learning_rate -- the learning rate, scalar.
Returns:
parameters -- python dictionary containing your updated parameters
"""
L = len(parameters) // 2 # number of layers in the neural networks
# Update rule for each parameter
for l in range(L):
### START CODE HERE ### (approx. 2 lines)
parameters["W" + str(l+1)] = parameters['W'+str(l+1)] - learning_rate * grads['dW'+str(l+1)]
parameters["b" + str(l+1)] = parameters['b'+str(l+1)] - learning_rate * grads['db'+str(l+1)]
### END CODE HERE ###
return parameters
(Batch) Gradient Descent:
X = data_input
Y = labels
parameters = initialize_parameters(layers_dims)
for i in range(0, num_iterations):
# Forward propagation
a, caches = forward_propagation(X, parameters)
# Compute cost.
cost = compute_cost(a, Y)
# Backward propagation.
grads = backward_propagation(a, caches, parameters)
# Update parameters.
parameters = update_parameters(parameters, grads)
Stochastic Gradient Descent:
X = data_input
Y = labels
parameters = initialize_parameters(layers_dims)
for i in range(0, num_iterations):
for j in range(0, m):
# Forward propagation
a, caches = forward_propagation(X[:,j], parameters)
# Compute cost
cost = compute_cost(a, Y[:,j])
# Backward propagation
grads = backward_propagation(a, caches, parameters)
# Update parameters.
parameters = update_parameters(parameters, grads)
In Stochastic Gradient Descent, you use only 1 training example before updating the gradients. When the training set is large, SGD can be faster. But the parameters will "oscillate" toward the minimum rather than converge smoothly.
Note also that implementing SGD requires 3 for-loops in total:
Over the number of iterations
Over the $m$ training examples
Over the layers (to update all parameters, from $(W^{[1]},b^{[1]})$ to $(W^{[L]},b^{[L]})$)
In practice, you'll often get faster results if you do not use neither the whole training set, nor only one training example, to perform each update. Mini-batch gradient descent uses an intermediate number of examples for each step. With mini-batch gradient descent, you loop over the mini-batches instead of looping over individual training examples.
What you should remember:
The difference between gradient descent, mini-batch gradient descent and stochastic gradient descent is the number of examples you use to perform one update step.
You have to tune a learning rate hyperparameter $\alpha$.
With a well-turned mini-batch size, usually it outperforms either gradient descent or stochastic gradient descent (particularly when the training set is large).
2 - Mini-Batch Gradient descent
Let's learn how to build mini-batches from the training set (X, Y).
There are two steps:
Shuffle: Create a shuffled version of the training set (X, Y) as shown below. Each column of X and Y represents a training example. Note that the random shuffling is done synchronously between X and Y. Such that after the shuffling the $i^{th}$ column of X is the example corresponding to the $i^{th}$ label in Y. The shuffling step ensures that examples will be split randomly into different mini-batches.
Partition: Partition the shuffled (X, Y) into mini-batches of size mini_batch_size (here 64). Note that the number of training examples is not always divisible by mini_batch_size. The last mini batch might be smaller, but you don't need to worry about this. When the final mini-batch is smaller than the full mini_batch_size, it will look like this:
selects the indexes for the $1^{st}$ and $2^{nd}$ mini-batches:
Note that the last mini-batch might end up smaller than mini_batch_size=64. Let $\lfloor s \rfloor$ represents $s$ rounded down to the nearest integer (this is math.floor(s) in Python). If the total number of examples is not a multiple of mini_batch_size=64 then there will be $\lfloor \frac{m}{mini_batch_size}\rfloor$ mini-batches with a full 64 examples, and the number of examples in the final mini-batch will be ($m-mini__batch__size \times \lfloor \frac{m}{mini_batch_size}\rfloor$).
def random_mini_batches(X, Y, mini_batch_size = 64, seed = 0):
"""
Creates a list of random minibatches from (X, Y)
Arguments:
X -- input data, of shape (input size, number of examples)
Y -- true "label" vector (1 for blue dot / 0 for red dot), of shape (1, number of examples)
mini_batch_size -- size of the mini-batches, integer
Returns:
mini_batches -- list of synchronous (mini_batch_X, mini_batch_Y)
"""
np.random.seed(seed) # To make your "random" minibatches the same as ours
m = X.shape[1] # number of training examples
mini_batches = []
# Step 1: Shuffle (X, Y)
permutation = list(np.random.permutation(m))
shuffled_X = X[:, permutation]
shuffled_Y = Y[:, permutation].reshape((1,m))
# Step 2: Partition (shuffled_X, shuffled_Y). Minus the end case.
num_complete_minibatches = math.floor(m/mini_batch_size) # number of mini batches of size mini_batch_size in your partitionning
for k in range(0, num_complete_minibatches):
### START CODE HERE ### (approx. 2 lines)
mini_batch_X = shuffled_X[:, k*mini_batch_size:(k+1)*mini_batch_size]
mini_batch_Y = shuffled_Y[:, k*mini_batch_size:(k+1)*mini_batch_size]
### END CODE HERE ###
mini_batch = (mini_batch_X, mini_batch_Y)
mini_batches.append(mini_batch)
# Handling the end case (last mini-batch < mini_batch_size)
if m % mini_batch_size != 0:
### START CODE HERE ### (approx. 2 lines)
mini_batch_X = shuffled_X[:, num_complete_minibatches*mini_batch_size:]
mini_batch_Y = shuffled_Y[:, num_complete_minibatches*mini_batch_size:]
### END CODE HERE ###
mini_batch = (mini_batch_X, mini_batch_Y)
mini_batches.append(mini_batch)
return mini_batches
Note:Powers of two are often chosen to be the mini-batch size, e.g., 16, 32, 64, 128.
3 - Momentum
Because mini-batch gradient descent makes a parameter update after seeing just a subset of examples, the direction of the update has some variance, and so the path taken by mini-batch gradient descent will "oscillate" toward convergence. Using momentum can reduce these oscillations.
Momentum takes into account the past gradients to smooth out the update. We will store the 'direction' of the previous gradients in the variable $v$. Formally, this will be the exponentially weighted average of the gradient on previous steps. You can also think of $v$ as the "velocity" of a ball rolling downhill, building up speed (and momentum) according to the direction of the gradient/slope of the hill.
Exercise: Initialize the velocity. The velocity, $v$, is a python dictionary that needs to be initialized with arrays of zeros. Its keys are the same as those in the grads dictionary, that is: for $l =1,...,L$:
v["dW" + str(l+1)] = ... #(numpy array of zeros with the same shape as parameters["W" + str(l+1)])
v["db" + str(l+1)] = ... #(numpy array of zeros with the same shape as parameters["b" + str(l+1)])
Note that the iterator l starts at 0 in the for loop while the first parameters are v["dW1"] and v["db1"] (that's a "one" on the superscript). This is why we are shifting l to l+1 in the for loop.
def initialize_velocity(parameters):
"""
Initializes the velocity as a python dictionary with:
- keys: "dW1", "db1", ..., "dWL", "dbL"
- values: numpy arrays of zeros of the same shape as the corresponding gradients/parameters.
Arguments:
parameters -- python dictionary containing your parameters.
parameters['W' + str(l)] = Wl
parameters['b' + str(l)] = bl
Returns:
v -- python dictionary containing the current velocity.
v['dW' + str(l)] = velocity of dWl
v['db' + str(l)] = velocity of dbl
"""
L = len(parameters) // 2 # number of layers in the neural networks
v = {}
# Initialize velocity
for l in range(L):
### START CODE HERE ### (approx. 2 lines)
v["dW" + str(l+1)] = np.zeros_like(parameters['W'+str(l+1)])
v["db" + str(l+1)] = np.zeros_like(parameters['b'+str(l+1)])
### END CODE HERE ###
return v
Exercise: Now, implement the parameters update with momentum. The momentum update rule is, for $l = 1, ..., L$:
where L is the number of layers, $\beta$ is the momentum and $\alpha$ is the learning rate. All parameters should be stored in the parameters dictionary. Note that the iterator l starts at 0 in the for loop while the first parameters are $W^{[1]}$ and $b^{[1]}$ (that's a "one" on the superscript). So you will need to shift l to l+1 when coding.
def update_parameters_with_momentum(parameters, grads, v, beta, learning_rate):
"""
Update parameters using Momentum
Arguments:
parameters -- python dictionary containing your parameters:
parameters['W' + str(l)] = Wl
parameters['b' + str(l)] = bl
grads -- python dictionary containing your gradients for each parameters:
grads['dW' + str(l)] = dWl
grads['db' + str(l)] = dbl
v -- python dictionary containing the current velocity:
v['dW' + str(l)] = ...
v['db' + str(l)] = ...
beta -- the momentum hyperparameter, scalar
learning_rate -- the learning rate, scalar
Returns:
parameters -- python dictionary containing your updated parameters
v -- python dictionary containing your updated velocities
"""
L = len(parameters) // 2 # number of layers in the neural networks
# Momentum update for each parameter
for l in range(L):
### START CODE HERE ### (approx. 4 lines)
# compute velocities
v["dW" + str(l+1)] = beta * v["dW" + str(l+1)] + (1 - beta) * grads['dW'+str(l+1)]
v["db" + str(l+1)] = beta * v["db" + str(l+1)] + (1 - beta) * grads['db'+str(l+1)]
# update parameters
parameters["W" + str(l+1)] = parameters["W" + str(l+1)] - learning_rate * v["dW" + str(l+1)]
parameters["b" + str(l+1)] = parameters["b" + str(l+1)] - learning_rate * v["db" + str(l+1)]
### END CODE HERE ###
return parameters, v
Note that:
The velocity is initialized with zeros. So the algorithm will take a few iterations to "build up" velocity and start to take bigger steps.
If $\beta = 0$, then this just becomes standard gradient descent without momentum.
How do you choose $\beta$?
The larger the momentum $\beta$ is, the smoother the update because the more we take the past gradients into account. But if $\beta$ is too big, it could also smooth out the updates too much.
Common values for $\beta$ range from 0.8 to 0.999. If you don't feel inclined to tune this, $\beta = 0.9$ is often a reasonable default.
Tuning the optimal $\beta$ for your model might need trying several values to see what works best in term of reducing the value of the cost function $J$.
What you should remember:
Momentum takes past gradients into account to smooth out the steps of gradient descent. It can be applied with batch gradient descent, mini-batch gradient descent or stochastic gradient descent.
You have to tune a momentum hyperparameter $\beta$ and a learning rate $\alpha$.
4 - Adam
Adam is one of the most effective optimization algorithms for training neural networks. It combines ideas from RMSProp (described in lecture) and Momentum.
How does Adam work?
It calculates an exponentially weighted average of past gradients, and stores it in variables $v$ (before bias correction) and $v^{corrected}$ (with bias correction).
It calculates an exponentially weighted average of the squares of the past gradients, and stores it in variables $s$ (before bias correction) and $s^{corrected}$ (with bias correction).
It updates parameters in a direction based on combining information from "1" and "2".
$\beta_1$ and $\beta_2$ are hyperparameters that control the two exponentially weighted averages.
$\alpha$ is the learning rate
$\varepsilon$ is a very small number to avoid dividing by zero
As usual, we will store all parameters in the parameters dictionary
Exercise: Initialize the Adam variables $v, s$ which keep track of the past information.
Instruction: The variables $v, s$ are python dictionaries that need to be initialized with arrays of zeros. Their keys are the same as for grads, that is: for $l = 1, ..., L$:
v["dW" + str(l+1)] = ... #(numpy array of zeros with the same shape as parameters["W" + str(l+1)])
v["db" + str(l+1)] = ... #(numpy array of zeros with the same shape as parameters["b" + str(l+1)])
s["dW" + str(l+1)] = ... #(numpy array of zeros with the same shape as parameters["W" + str(l+1)])
s["db" + str(l+1)] = ... #(numpy array of zeros with the same shape as parameters["b" + str(l+1)])
def initialize_adam(parameters) :
"""
Initializes v and s as two python dictionaries with:
- keys: "dW1", "db1", ..., "dWL", "dbL"
- values: numpy arrays of zeros of the same shape as the corresponding gradients/parameters.
Arguments:
parameters -- python dictionary containing your parameters.
parameters["W" + str(l)] = Wl
parameters["b" + str(l)] = bl
Returns:
v -- python dictionary that will contain the exponentially weighted average of the gradient.
v["dW" + str(l)] = ...
v["db" + str(l)] = ...
s -- python dictionary that will contain the exponentially weighted average of the squared gradient.
s["dW" + str(l)] = ...
s["db" + str(l)] = ...
"""
L = len(parameters) // 2 # number of layers in the neural networks
v = {}
s = {}
# Initialize v, s. Input: "parameters". Outputs: "v, s".
for l in range(L):
### START CODE HERE ### (approx. 4 lines)
v["dW" + str(l+1)] = np.zeros_like(parameters['W'+str(l+1)])
v["db" + str(l+1)] = np.zeros_like(parameters['b'+str(l+1)])
s["dW" + str(l+1)] = np.zeros_like(parameters['W'+str(l+1)])
s["db" + str(l+1)] = np.zeros_like(parameters['b'+str(l+1)])
### END CODE HERE ###
return v, s
Now, implement the parameters update with Adam.
def update_parameters_with_adam(parameters, grads, v, s, t, learning_rate = 0.01,
beta1 = 0.9, beta2 = 0.999, epsilon = 1e-8):
"""
Update parameters using Adam
Arguments:
parameters -- python dictionary containing your parameters:
parameters['W' + str(l)] = Wl
parameters['b' + str(l)] = bl
grads -- python dictionary containing your gradients for each parameters:
grads['dW' + str(l)] = dWl
grads['db' + str(l)] = dbl
v -- Adam variable, moving average of the first gradient, python dictionary
s -- Adam variable, moving average of the squared gradient, python dictionary
learning_rate -- the learning rate, scalar.
beta1 -- Exponential decay hyperparameter for the first moment estimates
beta2 -- Exponential decay hyperparameter for the second moment estimates
epsilon -- hyperparameter preventing division by zero in Adam updates
Returns:
parameters -- python dictionary containing your updated parameters
v -- Adam variable, moving average of the first gradient, python dictionary
s -- Adam variable, moving average of the squared gradient, python dictionary
"""
L = len(parameters) // 2 # number of layers in the neural networks
v_corrected = {} # Initializing first moment estimate, python dictionary
s_corrected = {} # Initializing second moment estimate, python dictionary
# Perform Adam update on all parameters
for l in range(L):
# Moving average of the gradients. Inputs: "v, grads, beta1". Output: "v".
### START CODE HERE ### (approx. 2 lines)
v["dW" + str(l+1)] = beta1 * v["dW" + str(l+1)] + (1 - beta1) * grads['dW'+str(l+1)]
v["db" + str(l+1)] = beta1 * v["db" + str(l+1)] + (1 - beta1) * grads['db'+str(l+1)]
### END CODE HERE ###
# Compute bias-corrected first moment estimate. Inputs: "v, beta1, t". Output: "v_corrected".
### START CODE HERE ### (approx. 2 lines)
v_corrected["dW" + str(l+1)] = v["dW" + str(l+1)] / (1 - beta1**t)
v_corrected["db" + str(l+1)] = v["db" + str(l+1)] / (1 - beta1**t)
# v_corrected["dW" + str(l+1)] = v["dW" + str(l+1)]/(1 - pow(beta1, t))
# v_corrected["db" + str(l+1)] = v["db" + str(l+1)]/(1 - pow(beta1, t))
### END CODE HERE ###
# Moving average of the squared gradients. Inputs: "s, grads, beta2". Output: "s".
### START CODE HERE ### (approx. 2 lines)
s["dW" + str(l+1)] = beta2 * s["dW" + str(l+1)] + (1 - beta2) * (grads["dW" + str(l+1)])**2
s["db" + str(l+1)] = beta2 * s["db" + str(l+1)] + (1 - beta2) * (grads["db" + str(l+1)])**2
# s["dW" + str(l+1)] = beta2 * s["dW" + str(l+1)] + (1 - beta2) * np.square(grads['dW'+str(l+1)])
# s["db" + str(l+1)] = beta2 * s["db" + str(l+1)] + (1 - beta2) * np.square(grads['db'+str(l+1)])
### END CODE HERE ###
# Compute bias-corrected second raw moment estimate. Inputs: "s, beta2, t". Output: "s_corrected".
### START CODE HERE ### (approx. 2 lines)
s_corrected["dW" + str(l+1)] = s["dW" + str(l+1)] / (1 - beta2**t)
s_corrected["db" + str(l+1)] = s["db" + str(l+1)] / (1 - beta2**t)
# s_corrected["dW" + str(l+1)] = s["dW" + str(l+1)]/(1 - pow(beta2, t))
# s_corrected["db" + str(l+1)] = s["db" + str(l+1)]/(1 - pow(beta2, t))
### END CODE HERE ###
# Update parameters. Inputs: "parameters, learning_rate, v_corrected, s_corrected, epsilon". Output: "parameters".
### START CODE HERE ### (approx. 2 lines)
parameters["W" + str(l+1)] = parameters["W" + str(l+1)] - learning_rate * v_corrected["dW" + str(l+1)] / (np.sqrt(s_corrected["dW" + str(l+1)]) + epsilon)
parameters["b" + str(l+1)] = parameters["b" + str(l+1)] - learning_rate * v_corrected["db" + str(l+1)] / (np.sqrt(s_corrected["db" + str(l+1)]) + epsilon)
### END CODE HERE ###
return parameters, v, s
5 - Model with different optimization algorithms
def model(X, Y, layers_dims, optimizer, learning_rate = 0.0007, mini_batch_size = 64, beta = 0.9,
beta1 = 0.9, beta2 = 0.999, epsilon = 1e-8, num_epochs = 10000, print_cost = True):
"""
3-layer neural network model which can be run in different optimizer modes.
Arguments:
X -- input data, of shape (2, number of examples)
Y -- true "label" vector (1 for blue dot / 0 for red dot), of shape (1, number of examples)
layers_dims -- python list, containing the size of each layer
learning_rate -- the learning rate, scalar.
mini_batch_size -- the size of a mini batch
beta -- Momentum hyperparameter
beta1 -- Exponential decay hyperparameter for the past gradients estimates
beta2 -- Exponential decay hyperparameter for the past squared gradients estimates
epsilon -- hyperparameter preventing division by zero in Adam updates
num_epochs -- number of epochs
print_cost -- True to print the cost every 1000 epochs
Returns:
parameters -- python dictionary containing your updated parameters
"""
L = len(layers_dims) # number of layers in the neural networks
costs = [] # to keep track of the cost
t = 0 # initializing the counter required for Adam update
seed = 10 # For grading purposes, so that your "random" minibatches are the same as ours
# Initialize parameters
parameters = initialize_parameters(layers_dims)
# Initialize the optimizer
if optimizer == "gd":
pass # no initialization required for gradient descent
elif optimizer == "momentum":
v = initialize_velocity(parameters)
elif optimizer == "adam":
v, s = initialize_adam(parameters)
# Optimization loop
for i in range(num_epochs):
# Define the random minibatches. We increment the seed to reshuffle differently the dataset after each epoch
seed = seed + 1
minibatches = random_mini_batches(X, Y, mini_batch_size, seed)
for minibatch in minibatches:
# Select a minibatch
(minibatch_X, minibatch_Y) = minibatch
# Forward propagation
a3, caches = forward_propagation(minibatch_X, parameters)
# Compute cost
cost = compute_cost(a3, minibatch_Y)
# Backward propagation
grads = backward_propagation(minibatch_X, minibatch_Y, caches)
# Update parameters
if optimizer == "gd":
parameters = update_parameters_with_gd(parameters, grads, learning_rate)
elif optimizer == "momentum":
parameters, v = update_parameters_with_momentum(parameters, grads, v, beta, learning_rate)
elif optimizer == "adam":
t = t + 1 # Adam counter
parameters, v, s = update_parameters_with_adam(parameters, grads, v, s,
t, learning_rate, beta1, beta2, epsilon)
# Print the cost every 1000 epoch
if print_cost and i % 1000 == 0:
print ("Cost after epoch %i: %f" %(i, cost))
if print_cost and i % 100 == 0:
costs.append(cost)
# plot the cost
plt.plot(costs)
plt.ylabel('cost')
plt.xlabel('epochs (per 100)')
plt.title("Learning rate = " + str(learning_rate))
plt.show()
return parameters
summary
Momentum usually helps, but given the small learning rate and the simplistic dataset, its impact is almost negligeable. Also, the huge oscillations you see in the cost come from the fact that some minibatches are more difficult thans others for the optimization algorithm.
Adam on the other hand, clearly outperforms mini-batch gradient descent and Momentum. If you run the model for more epochs on this simple dataset, all three methods will lead to very good results. However, you've seen that Adam converges a lot faster.
Some advantages of Adam include:
Relatively low memory requirements (though higher than gradient descent and gradient descent with momentum)
Usually works well even with little tuning of hyperparameters (except $\alpha$)
TensorFlow
1 - Exploring the Tensorflow Library
loss = \mathcal{L}(\hat{y}, y) = (\hat y^{(i)} - y^{(i)})^2 \tag{1}
import tensorflow as tf
from tensorflow.python.framework import ops
y_hat = tf.constant(36, name='y_hat') # Define y_hat constant. Set to 36.
y = tf.constant(39, name='y') # Define y. Set to 39
loss = tf.Variable((y - y_hat)**2, name='loss') # Create a variable for the loss
init = tf.global_variables_initializer() # When init is run later (session.run(init)),
# the loss variable will be initialized and ready to be computed
with tf.Session() as session: # Create a session and print the output
session.run(init) # Initializes the variables
print(session.run(loss)) # Prints the loss
Writing and running programs in TensorFlow has the following steps:
Create Tensors (variables) that are not yet executed/evaluated.
Write operations between those Tensors.
Initialize your Tensors.
Create a Session.
Run the Session. This will run the operations you'd written above.
Therefore, when we created a variable for the loss, we simply defined the loss as a function of other quantities, but did not evaluate its value. To evaluate it, we had to run init=tf.global_variables_initializer(). That initialized the loss variable, and in the last line we were finally able to evaluate the value of loss and print its value.
remember to initialize your variables, create a session and run the operations inside the session.
Next, you'll also have to know about placeholders. A placeholder is an object whose value you can specify only later. To specify values for a placeholder, you can pass in values by using a "feed dictionary" (feed_dict variable). Below, we created a placeholder for x. This allows us to pass in a number later when we run the session.
x = tf.placeholder(tf.int64, name = 'x')
print(sess.run(2 * x, feed_dict = {x: 3}))
sess.close()
When you first defined x you did not have to specify a value for it. A placeholder is simply a variable that you will assign data to only later, when running the session. We say that you feed data to these placeholders when running the session.
Here's what's happening: When you specify the operations needed for a computation, you are telling TensorFlow how to construct a computation graph. The computation graph can have some placeholders whose values you will specify only later. Finally, when you run the session, you are telling TensorFlow to execute the computation graph.
1.1 - Linear function
def linear_function():
"""
Implements a linear function:
Initializes W to be a random tensor of shape (4,3)
Initializes X to be a random tensor of shape (3,1)
Initializes b to be a random tensor of shape (4,1)
Returns:
result -- runs the session for Y = WX + b
"""
np.random.seed(1)
### START CODE HERE ### (4 lines of code)
X = tf.constant(np.random.randn(3,1), name = "X")
W = tf.constant(np.random.randn(4,3), name = "W")
b = tf.constant(np.random.randn(4,1), name = "b")
Y = tf.add(tf.matmul(W,X), b)
### END CODE HERE ###
# Create the session using tf.Session() and run it with sess.run(...) on the variable you want to calculate
### START CODE HERE ###
sess = tf.Session()
result = sess.run(Y)
### END CODE HERE ###
# close the session
sess.close()
return result
1.2 - Computing the sigmoid
def sigmoid(z):
"""
Computes the sigmoid of z
Arguments:
z -- input value, scalar or vector
Returns:
results -- the sigmoid of z
"""
### START CODE HERE ### ( approx. 4 lines of code)
# Create a placeholder for x. Name it 'x'.
x = tf.placeholder(tf.float32, name="x")
# compute sigmoid(x)
sigmoid = tf.sigmoid(x)
# Create a session, and run it. Please use the method 2 explained above.
# You should use a feed_dict to pass z's value to x.
with tf.Session() as sess:
# Run session and call the output "result"
result = sess.run(sigmoid, feed_dict={x:z})
### END CODE HERE ###
return result
1.3 - Computing the Cost
def cost(logits, labels):
"""
Computes the cost using the sigmoid cross entropy
Arguments:
logits -- vector containing z, output of the last linear unit (before the final sigmoid activation)
labels -- vector of labels y (1 or 0)
Note: What we've been calling "z" and "y" in this class are respectively called "logits" and "labels"
in the TensorFlow documentation. So logits will feed into z, and labels into y.
Returns:
cost -- runs the session of the cost (formula (2))
"""
### START CODE HERE ###
# Create the placeholders for "logits" (z) and "labels" (y) (approx. 2 lines)
z = tf.placeholder(tf.float32, name="logits")
y = tf.placeholder(tf.float32, name="labels")
# Use the loss function (approx. 1 line)
cost = tf.nn.sigmoid_cross_entropy_with_logits(logits=z, labels=y)
# Create a session (approx. 1 line). See method 1 above.
sess = tf.Session()
# Run the session (approx. 1 line).
cost = sess.run(cost, feed_dict={z:logits, y:labels})
# Close the session (approx. 1 line). See method 1 above.
sess.close()
### END CODE HERE ###
return cost
1.4 - Using One Hot encodings
def one_hot_matrix(labels, C):
"""
Creates a matrix where the i-th row corresponds to the ith class number and the jth column
corresponds to the jth training example. So if example j had a label i. Then entry (i,j)
will be 1.
Arguments:
labels -- vector containing the labels
C -- number of classes, the depth of the one hot dimension
Returns:
one_hot -- one hot matrix
"""
### START CODE HERE ###
# Create a tf.constant equal to C (depth), name it 'C'. (approx. 1 line)
C = tf.constant(C, name='C')
# Use tf.one_hot, be careful with the axis (approx. 1 line)
one_hot_matrix = tf.one_hot(labels, C, axis=0)
# Create the session (approx. 1 line)
sess = tf.Session()
# Run the session (approx. 1 line)
one_hot = sess.run(one_hot_matrix)
# Close the session (approx. 1 line). See method 1 above.
sess.close()
### END CODE HERE ###
return one_hot
1.5 - Initialize with zeros and ones
def ones(shape):
"""
Creates an array of ones of dimension shape
Arguments:
shape -- shape of the array you want to create
Returns:
ones -- array containing only ones
"""
### START CODE HERE ###
# Create "ones" tensor using tf.ones(...). (approx. 1 line)
ones = tf.ones(shape)
# Create the session (approx. 1 line)
sess = tf.Session()
# Run the session to compute 'ones' (approx. 1 line)
ones = sess.run(ones)
# Close the session (approx. 1 line). See method 1 above.
sess.close()
### END CODE HERE ###
return ones
##2 - Building your first neural network in tensorflow
2.1 - Create placeholders
def create_placeholders(n_x, n_y):
"""
Creates the placeholders for the tensorflow session.
Arguments:
n_x -- scalar, size of an image vector (num_px * num_px = 64 * 64 * 3 = 12288)
n_y -- scalar, number of classes (from 0 to 5, so -> 6)
Returns:
X -- placeholder for the data input, of shape [n_x, None] and dtype "float"
Y -- placeholder for the input labels, of shape [n_y, None] and dtype "float"
Tips:
- You will use None because it let's us be flexible on the number of examples you will for the placeholders.
In fact, the number of examples during test/train is different.
"""
### START CODE HERE ### (approx. 2 lines)
X = tf.placeholder(tf.float32, shape=(n_x,None), name="X")
Y = tf.placeholder(tf.float32, shape=(n_y,None), name="Y")
### END CODE HERE ###
return X, Y
Question: Implement the forward pass of the neural network. We commented for you the numpy equivalents so that you can compare the tensorflow implementation to numpy. It is important to note that the forward propagation stops at z3. The reason is that in tensorflow the last linear layer output is given as input to the function computing the loss. Therefore, you don't need a3!
def forward_propagation(X, parameters):
"""
Implements the forward propagation for the model: LINEAR -> RELU -> LINEAR -> RELU -> LINEAR -> SOFTMAX
Arguments:
X -- input dataset placeholder, of shape (input size, number of examples)
parameters -- python dictionary containing your parameters "W1", "b1", "W2", "b2", "W3", "b3"
the shapes are given in initialize_parameters
Returns:
Z3 -- the output of the last LINEAR unit
"""
# Retrieve the parameters from the dictionary "parameters"
W1 = parameters['W1']
b1 = parameters['b1']
W2 = parameters['W2']
b2 = parameters['b2']
W3 = parameters['W3']
b3 = parameters['b3']
### START CODE HERE ### (approx. 5 lines) # Numpy Equivalents:
Z1 = tf.matmul(W1, X)+b1 # Z1 = np.dot(W1, X) + b1
A1 = tf.nn.relu(Z1) # A1 = relu(Z1)
Z2 = tf.matmul(W2, A1)+b2 # Z2 = np.dot(W2, a1) + b2
A2 = tf.nn.relu(Z2) # A2 = relu(Z2)
Z3 = tf.matmul(W3, A2)+b3 # Z3 = np.dot(W3,Z2) + b3
### END CODE HERE ###
return Z3
2.4 Compute cost
Question: Implement the cost function below.
It is important to know that the "logits" and "labels" inputs of tf.nn.softmax_cross_entropy_with_logits are expected to be of shape (number of examples, num_classes). We have thus transposed Z3 and Y for you.
Besides, tf.reduce_mean basically does the summation over the examples.
def compute_cost(Z3, Y):
"""
Computes the cost
Arguments:
Z3 -- output of forward propagation (output of the last LINEAR unit), of shape (6, number of examples)
Y -- "true" labels vector placeholder, same shape as Z3
Returns:
cost - Tensor of the cost function
"""
# to fit the tensorflow requirement for tf.nn.softmax_cross_entropy_with_logits(...,...)
logits = tf.transpose(Z3)
labels = tf.transpose(Y)
### START CODE HERE ### (1 line of code)
cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(logits = logits, labels = labels))
### END CODE HERE ###
return cost
2.5 - Backward propagation & parameter updates
All the backpropagation and the parameters update is taken care of in 1 line of code. It is very easy to incorporate this line in the model.
After you compute the cost function. You will create an "optimizer" object. You have to call this object along with the cost when running the tf.session. When called, it will perform an optimization on the given cost with the chosen method and learning rate.
For instance, for gradient descent the optimizer would be:
_ , c = sess.run([optimizer, cost], feed_dict={X: minibatch_X, Y: minibatch_Y})
This computes the backpropagation by passing through the tensorflow graph in the reverse order. From cost to inputs.
Note When coding, we often use _ as a "throwaway" variable to store values that we won't need to use later. Here, _ takes on the evaluated value of optimizer, which we don't need (and c takes the value of the cost variable).
2.6 - Building the model
def model(X_train, Y_train, X_test, Y_test, learning_rate = 0.0001,
num_epochs = 1500, minibatch_size = 32, print_cost = True):
"""
Implements a three-layer tensorflow neural network: LINEAR->RELU->LINEAR->RELU->LINEAR->SOFTMAX.
Arguments:
X_train -- training set, of shape (input size = 12288, number of training examples = 1080)
Y_train -- test set, of shape (output size = 6, number of training examples = 1080)
X_test -- training set, of shape (input size = 12288, number of training examples = 120)
Y_test -- test set, of shape (output size = 6, number of test examples = 120)
learning_rate -- learning rate of the optimization
num_epochs -- number of epochs of the optimization loop
minibatch_size -- size of a minibatch
print_cost -- True to print the cost every 100 epochs
Returns:
parameters -- parameters learnt by the model. They can then be used to predict.
"""
ops.reset_default_graph() # to be able to rerun the model without overwriting tf variables
tf.set_random_seed(1) # to keep consistent results
seed = 3 # to keep consistent results
(n_x, m) = X_train.shape # (n_x: input size, m : number of examples in the train set)
n_y = Y_train.shape[0] # n_y : output size
costs = [] # To keep track of the cost
# Create Placeholders of shape (n_x, n_y)
### START CODE HERE ### (1 line)
X, Y = create_placeholders(n_x, n_y)
### END CODE HERE ###
# Initialize parameters
### START CODE HERE ### (1 line)
parameters = initialize_parameters()
### END CODE HERE ###
# Forward propagation: Build the forward propagation in the tensorflow graph
### START CODE HERE ### (1 line)
Z3 = forward_propagation(X, parameters)
# Z3 = tf.nn.sigmoid(Z3)
# Z3 = tf.nn.softmax(Z3)
### END CODE HERE ###
# Cost function: Add cost function to tensorflow graph
### START CODE HERE ### (1 line)
cost = compute_cost(Z3, Y)
### END CODE HERE ###
# Backpropagation: Define the tensorflow optimizer. Use an AdamOptimizer.
### START CODE HERE ### (1 line)
optimizer = tf.train.AdamOptimizer(learning_rate).minimize(cost)
# optimizer = tf.train.GradientDescentOptimizer(learning_rate).minimize(cost)
### END CODE HERE ###
# Initialize all the variables
init = tf.global_variables_initializer()
# Start the session to compute the tensorflow graph
with tf.Session() as sess:
# Run the initialization
sess.run(init)
# Do the training loop
for epoch in range(num_epochs):
epoch_cost = 0. # Defines a cost related to an epoch
num_minibatches = int(m / minibatch_size) # number of minibatches of size minibatch_size in the train set
seed = seed + 1
minibatches = random_mini_batches(X_train, Y_train, minibatch_size, seed)
for minibatch in minibatches:
# Select a minibatch
(minibatch_X, minibatch_Y) = minibatch
# IMPORTANT: The line that runs the graph on a minibatch.
# Run the session to execute the "optimizer" and the "cost", the feedict should contain a minibatch for (X,Y).
### START CODE HERE ### (1 line)
_ , minibatch_cost = sess.run([optimizer, cost], feed_dict={X: minibatch_X, Y: minibatch_Y})
### END CODE HERE ###
epoch_cost += minibatch_cost / num_minibatches
# Print the cost every epoch
if print_cost == True and epoch % 100 == 0:
print ("Cost after epoch %i: %f" % (epoch, epoch_cost))
if print_cost == True and epoch % 5 == 0:
costs.append(epoch_cost)
# plot the cost
plt.plot(np.squeeze(costs))
plt.ylabel('cost')
plt.xlabel('iterations (per tens)')
plt.title("Learning rate =" + str(learning_rate))
plt.show()
# lets save the parameters in a variable
parameters = sess.run(parameters)
print ("Parameters have been trained!")
# Calculate the correct predictions
correct_prediction = tf.equal(tf.argmax(Z3), tf.argmax(Y))
# Calculate accuracy on the test set
accuracy = tf.reduce_mean(tf.cast(correct_prediction, "float"))
print ("Train Accuracy:", accuracy.eval({X: X_train, Y: Y_train}))
print ("Test Accuracy:", accuracy.eval({X: X_test, Y: Y_test}))
return parameters
###辅助函数
def random_mini_batches(X, Y, mini_batch_size = 64, seed = 0):
"""
Creates a list of random minibatches from (X, Y)
Arguments:
X -- input data, of shape (input size, number of examples)
Y -- true "label" vector (containing 0 if cat, 1 if non-cat), of shape (1, number of examples)
mini_batch_size - size of the mini-batches, integer
seed -- this is only for the purpose of grading, so that you're "random minibatches are the same as ours.
Returns:
mini_batches -- list of synchronous (mini_batch_X, mini_batch_Y)
"""
m = X.shape[1] # number of training examples
mini_batches = []
np.random.seed(seed)
# Step 1: Shuffle (X, Y)
permutation = list(np.random.permutation(m))
shuffled_X = X[:, permutation]
shuffled_Y = Y[:, permutation].reshape((Y.shape[0],m))
# Step 2: Partition (shuffled_X, shuffled_Y). Minus the end case.
num_complete_minibatches = math.floor(m/mini_batch_size) # number of mini batches of size mini_batch_size in your partitionning
for k in range(0, num_complete_minibatches):
mini_batch_X = shuffled_X[:, k * mini_batch_size : k * mini_batch_size + mini_batch_size]
mini_batch_Y = shuffled_Y[:, k * mini_batch_size : k * mini_batch_size + mini_batch_size]
mini_batch = (mini_batch_X, mini_batch_Y)
mini_batches.append(mini_batch)
# Handling the end case (last mini-batch < mini_batch_size)
if m % mini_batch_size != 0:
mini_batch_X = shuffled_X[:, num_complete_minibatches * mini_batch_size : m]
mini_batch_Y = shuffled_Y[:, num_complete_minibatches * mini_batch_size : m]
mini_batch = (mini_batch_X, mini_batch_Y)
mini_batches.append(mini_batch)
return mini_batches
def convert_to_one_hot(Y, C):
Y = np.eye(C)[Y.reshape(-1)].T
return Y
def predict(X, parameters):
W1 = tf.convert_to_tensor(parameters["W1"])
b1 = tf.convert_to_tensor(parameters["b1"])
W2 = tf.convert_to_tensor(parameters["W2"])
b2 = tf.convert_to_tensor(parameters["b2"])
W3 = tf.convert_to_tensor(parameters["W3"])
b3 = tf.convert_to_tensor(parameters["b3"])
params = {"W1": W1,
"b1": b1,
"W2": W2,
"b2": b2,
"W3": W3,
"b3": b3}
x = tf.placeholder("float", [12288, 1])
z3 = forward_propagation_for_predict(x, params)
p = tf.argmax(z3)
sess = tf.Session()
prediction = sess.run(p, feed_dict = {x: X})
return prediction
def forward_propagation_for_predict(X, parameters):
"""
Implements the forward propagation for the model: LINEAR -> RELU -> LINEAR -> RELU -> LINEAR -> SOFTMAX
Arguments:
X -- input dataset placeholder, of shape (input size, number of examples)
parameters -- python dictionary containing your parameters "W1", "b1", "W2", "b2", "W3", "b3"
the shapes are given in initialize_parameters
Returns:
Z3 -- the output of the last LINEAR unit
"""
# Retrieve the parameters from the dictionary "parameters"
W1 = parameters['W1']
b1 = parameters['b1']
W2 = parameters['W2']
b2 = parameters['b2']
W3 = parameters['W3']
b3 = parameters['b3']
# Numpy Equivalents:
Z1 = tf.add(tf.matmul(W1, X), b1) # Z1 = np.dot(W1, X) + b1
A1 = tf.nn.relu(Z1) # A1 = relu(Z1)
Z2 = tf.add(tf.matmul(W2, A1), b2) # Z2 = np.dot(W2, a1) + b2
A2 = tf.nn.relu(Z2) # A2 = relu(Z2)
Z3 = tf.add(tf.matmul(W3, A2), b3) # Z3 = np.dot(W3,Z2) + b3
return Z3
2.7 - Test with your own image
import scipy
from PIL import Image
from scipy import ndimage
## START CODE HERE ## (PUT YOUR IMAGE NAME)
my_image = "thumbs_up.jpg"
## END CODE HERE ##
# We preprocess your image to fit your algorithm.
fname = "images/" + my_image
image = np.array(ndimage.imread(fname, flatten=False))
my_image = scipy.misc.imresize(image, size=(64,64)).reshape((1, 64*64*3)).T
my_image_prediction = predict(my_image, parameters)
plt.imshow(image)
print("Your algorithm predicts: y = " + str(np.squeeze(my_image_prediction)))
Summary
What you should remember:
Tensorflow is a programming framework used in deep learning
The two main object classes in tensorflow are Tensors and Operators.
When you code in tensorflow you have to take the following steps:
Create a graph containing Tensors (Variables, Placeholders ...) and Operations (tf.matmul, tf.add, ...)
Create a session
Initialize the session
Run the session to execute the graph
You can execute the graph multiple times as you've seen in model()
The backpropagation and optimization is automatically done when running the session on the "optimizer" object.
Figure 3: The red arrows shows the direction taken by one step of mini-batch gradient descent with momentum. The blue points show the direction of the gradient (with respect to the current mini-batch) on each step. Rather than just following the gradient, we let the gradient influence $v$ and then take a step in the direction of $v$.
