09.10.2023 Views

Advanced Data Analytics Using Python_ With Machine Learning, Deep Learning and NLP Examples ( 2023)

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Chapter 5

Deep Learning and Neural Networks

to "self.biases" and "self.weights"."""

nabla_b1=[np.zeros(b1.shape)for b 1in self.biases]

nabla_w1=[np.zeros(w1.shape)for w1in self.weights]

# feedforward

activation1=x1

activations1=[x1]

zs1=[]

for b,winzip(self.biases,self.weights):

z1=np.dot(w1,activation1)+b1

zs1.append(z1)

activation1=sigmoid(z1)

activations1.append(activation1)

# backward pass

delta1=self.cost_derivative(activations1[-1],y1)* \

sigmoid_prime(zs1[-1])

nabla_b1[-1]=delta1

nabla_w1[-1]=np.dot(delta,activations1[-2].

transpose())

for l in xrange(2,self.num_layers):

z1=zs1[-l]

sp1=sigmoid_prime(z1)

delta1=np.dot(self.weights1[-l+1].

transpose(),delta)*sp1

nabla_b1[-l]=delta1

nabla_w1[-l]=np.

dot(delta,activations1[-l-1].transpose())

return(nabla_b1,nabla_w1)

def cost_derivative(self,output_activations,y):

"""Return the vector of partial derivatives \

partial C_x /

105

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!