my_answers.py

# python3
#
# PROGRAMMER: Félix Ramón López Martínez
# DATE CREATED: 5/11/2020
# REVISED DATE:
# PURPOSE: MPL neural network call from Jupyter Notebook
#          "Predicting_bike_sharing_data.ipynb"
#         Built as part of the Bike Sharing project of the Udacity Nanodegree
#         program Deep Learning
##

import numpy as np


class NeuralNetwork(object):
    def __init__(self, input_nodes, hidden_nodes, output_nodes, learning_rate):
        # Set number of nodes in input, hidden and output layers.
        self.input_nodes = input_nodes
        self.hidden_nodes = hidden_nodes
        self.output_nodes = output_nodes

        # Initialize weights
        self.weights_input_to_hidden = np.random.normal(0.0, self.input_nodes**-0.5,
                                       (self.input_nodes, self.hidden_nodes))

        self.weights_hidden_to_output = np.random.normal(0.0, self.hidden_nodes**-0.5,
                                       (self.hidden_nodes, self.output_nodes))
        self.lr = learning_rate

        #### DONE: Set self.activation_function to your implemented sigmoid function ####
        self.activation_function = lambda x : 1 / (1 + np.exp(-x))


    def train(self, features, targets):
        ''' Train the network on batch of features and targets.

            Arguments
            ---------

            features: 2D array, each row is one data record, each column is a feature
            targets: 1D array of target values

        '''
        n_records = features.shape[0]
        delta_weights_i_h = np.zeros(self.weights_input_to_hidden.shape)
        delta_weights_h_o = np.zeros(self.weights_hidden_to_output.shape)
        for X, y in zip(features, targets):

            # Call to FORWARD PASS function
            final_outputs, hidden_outputs = self.forward_pass_train(X)

            # Call to BACKPROPAGATION function
            delta_weights_i_h, delta_weights_h_o = self.backpropagation(final_outputs, hidden_outputs, X, y,
                                                                        delta_weights_i_h, delta_weights_h_o)
        self.update_weights(delta_weights_i_h, delta_weights_h_o, n_records)


    def forward_pass_train(self, X):
        ''' Implement forward pass here

            Arguments
            ---------
            X: features batch

        '''
        #### Implement the forward pass here ####
        ### Forward pass ###
        # DONE: Hidden layer - Replace these values with your calculations.
        hidden_inputs = np.dot(X, self.weights_input_to_hidden) # signals into hidden layer
        hidden_outputs = self.activation_function(hidden_inputs) # signals from hidden layer

        # DONE: Output layer - Replace these values with your calculations.
        final_inputs = np.dot(hidden_outputs, self.weights_hidden_to_output) # signals into final output layer
        final_outputs = final_inputs  # signals from final output layer (no activation function, f(x)=x

        return final_outputs, hidden_outputs

    def backpropagation(self, final_outputs, hidden_outputs, X, y, delta_weights_i_h, delta_weights_h_o):
        ''' Implement backpropagation

            Arguments
            ---------
            final_outputs: output from forward pass
            y: target (i.e. label) batch
            delta_weights_i_h: change in weights from input to hidden layers
            delta_weights_h_o: change in weights from hidden to output layers

        '''
        #### Implement the backward pass here ####
        ### Backward pass ###

        # DONE: Output error - Replace this value with your calculations.
        error = y - final_outputs # Output layer error is the difference between desired target and actual output.

        # DONE: Backpropagated error terms - Replace these values with your calculations.
        output_error_term = error  # activation function is identity function, so derivative is 1

        # DONE: Calculate the hidden layer's contribution to the error
        hidden_error = np.dot(output_error_term, self.weights_hidden_to_output.T)
        hidden_error_term = hidden_error * hidden_outputs * (1 - hidden_outputs)

        # Weight step (input to hidden)
        delta_weights_i_h += self.lr * np.dot(X[:, None], hidden_error_term[:, None].T)

        # Weight step (hidden to output)
        delta_weights_h_o += self.lr * np.dot(hidden_outputs[:, None], output_error_term[:, None].T)

        return delta_weights_i_h, delta_weights_h_o

    def update_weights(self, delta_weights_i_h, delta_weights_h_o, n_records):
        ''' Update weights on gradient descent step

            Arguments
            ---------
            delta_weights_i_h: change in weights from input to hidden layers
            delta_weights_h_o: change in weights from hidden to output layers
            n_records: number of records

        '''
        self.weights_hidden_to_output += delta_weights_h_o / n_records # update h2o weights with gradient descent step
        self.weights_input_to_hidden += delta_weights_i_h / n_records # update i2h weights with gradient descent step

    def run(self, features):
        ''' Run a forward pass through the network with input features

            Arguments
            ---------
            features: 1D array of feature values
        '''

        #### Implement the forward pass here ####

        #### Implement the forward pass here ####
        ### Forward pass ###
        # DONE: Hidden layer - Replace these values with your calculations.
        hidden_inputs = np.dot(features, self.weights_input_to_hidden) # signals into hidden layer
        hidden_outputs = self.activation_function(hidden_inputs) # signals from hidden layer

        # DONE: Output layer - Replace these values with your calculations.
        final_inputs = np.dot(hidden_outputs, self.weights_hidden_to_output) # signals into final output layer
        final_outputs = final_inputs  # signals from final output layer (no activation function, f(x)=x

        return final_outputs


#########################################################
# Set your hyperparameters here
##########################################################
iterations = 2000
learning_rate = 1.0
hidden_nodes = 11
output_nodes = 1