Scribd Logo
Upload

Welcome to Scribd!

  • 680 views

    DNN Handson

    Uploaded by

    Vinaykumarreddy
    This document defines functions for building and training a deep neural network model in TensorFlow. It initializes placeholders and parameters, defines forward propagation through linear and relu activation layers, calculates the cost using cross entropy, and trains the model using gradient descent. It runs the training for 10,000 iterations on sample data and returns the final parameters.

    Copyright:

    © All Rights Reserved
    Download as TXT, PDF, TXT or read online from Scribd

    DNN Handson

    Uploaded by

    Vinaykumarreddy
    680 views2 pages
    This document defines functions for building and training a deep neural network model in TensorFlow. It initializes placeholders and parameters, defines forward propagation through linear and relu activation layers, calculates the cost using cross entropy, and trains the model using gradient descent. It runs the training for 10,000 iterations on sample data and returns the final parameters.

    Original Description:

    dnn handson

    Original Title

    DNN Handson.txt

    Copyright

    © © All Rights Reserved

    Available Formats

    TXT, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    This document defines functions for building and training a deep neural network model in TensorFlow. It initializes placeholders and parameters, defines forward propagation through linear and relu activation layers, calculates the cost using cross entropy, and trains the model using gradient descent. It runs the training for 10,000 iterations on sample data and returns the final parameters.

    Copyright:

    © All Rights Reserved
    Download as TXT, PDF, TXT or read online from Scribd
    Download as txt, pdf, or txt
    680 views2 pages

    DNN Handson

    Uploaded by

    Vinaykumarreddy
    This document defines functions for building and training a deep neural network model in TensorFlow. It initializes placeholders and parameters, defines forward propagation through linear and relu activation layers, calculates the cost using cross entropy, and trains the model using gradient descent. It runs the training for 10,000 iterations on sample data and returns the final parameters.

    Copyright:

    © All Rights Reserved
    Download as TXT, PDF, TXT or read online from Scribd
    Download as txt, pdf, or txt
    of 2

    X_data =X.

    y_data =y.reshape(1,1000)

    layer_dims = [2,9,9,1]

    import tensorflow as tf

    A_0 = tf.placeholder(dtype = tf.float64, shape = ([num_features,None]))


    Y = tf.placeholder(dtype = tf.float64, shape = ([1,None]))

    def initialize_parameters_deep(layer_dims):
    L = len(layer_dims)
    parameters = {}
    for l in range(1,L):
    parameters['W' + str(l)] =
    tf.Variable(initial_value=tf.random_normal([layer_dims[l], layer_dims[l-1]],
    dtype=tf.float64)* 0.01)
    parameters['b' + str(l)] =
    tf.Variable(initial_value=tf.zeros([layer_dims[l],1],dtype=tf.float64) * 0.01)

    return parameters

    def linear_forward_prop(A_prev,W,b, activation):


    Z = tf.add(tf.matmul(W, A_prev), b)
    if activation == "sigmoid":
    A = Z
    elif activation == "relu":
    A = tf.nn.relu(Z) #apply relu
    activation on Z using tf.nn.relu() function
    return A

    def l_layer_forwardProp(A_0, parameters):


    A = A_0
    L = len(parameters)//2
    for l in range(1,L):
    A_prev = A
    A = linear_forward_prop(A_prev,parameters['W' + str(l)],parameters['b' +
    str(l)], "relu")
    A_final = linear_forward_prop(A, parameters['W' + str(L)], parameters['b' +
    str(L)], "sigmoid" )
    return A_final

    def deep_layer_network(X_data,y_data, layer_dims, learning_rate, num_iter):


    num_features = layer_dims[0]
    A_0, Y = placeholders(num_features) #call
    placeholder function to initialize placeholders A_0 and Y
    parameters = initialize_parameters_deep(layer_dims)
    #Initialse Weights and bias using initialize_parameters_deep() with layer_dims as
    parameters
    Z_final = l_layer_forwardProp(A_0,parameters)
    #call the function l_layer_forwardProp() to define the final output
    cost =
    tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(logits=Z_final,labels=Y))
    #define the mean cost using sigmoid cross entropy function
    train_net = tf.train.GradientDescentOptimizer(learning_rate).minimize(cost)
    #call tensorflow's gradient descent optimizer function with minimize cost
    init = tf.global_variables_initializer()
    with tf.Session() as sess:
    sess.run(init)
    for i in range(num_iter):
    _,c = sess.run([train_net, cost], feed_dict={A_0: X_data, Y: y_data})
    if i % 1000 == 0:
    print(c)
    with open("Output.txt", "w") as text_file:
    text_file.write("c= %f\n" % c)
    params = sess.run(parameters)
    return params

    parameters = deep_layer_network(X_data, y_data, layer_dims ,0.3, 10000)

    You might also like