Convolutional Neural Networks in Python: Beginner's Guide to Convolutional Neural Networks in Python

By Frank Millstein

This book covers the basics behind Convolutional Neural Networks by introducing you to this complex world of deep learning and artificial neural networks in a simple and easy to understand way. It is perfect for any beginner out there looking forward to learning more about this machine learning field.
This book is all about how to use convolutional neural networks for various image, object and other common classification problems in Python. Here, we also take a deeper look into various Keras layer used for building CNNs we take a look at different activation functions and much more, which will eventually lead you to creating highly accurate models able of performing great task results on various image classification, object classification and other problems.
Therefore, at the end of the book, you will have a better insight into this world, thus you will be more than prepared to deal with more complex and challenging tasks on your own.

Here Is a Preview of What You'll Learn In This Book…

• Convolutional neural networks structure
• How convolutional neural networks actually work
• Convolutional neural networks applications
• The importance of convolution operator
• Different convolutional neural networks layers and their importance
• Arrangement of spatial parameters
• How and when to use stride and zero-padding
• Method of parameter sharing
• Matrix multiplication and its importance
• Pooling and dense layers
• Introducing non-linearity relu activation function
• How to train your convolutional neural network models using backpropagation
• How and why to apply dropout
• CNN model training process
• How to build a convolutional neural network
• Generating predictions and calculating loss functions
• How to train and evaluate your MNIST classifier
• How to build a simple image classification CNN
• And much, much more!

Get this book NOW and learn more about Convolutional Neural Networks in Python!

• Language: English
• Publisher: Frank Millstein
• Release date: Oct 18, 2019
• ISBN: 9781393633716

Book preview

By Frank Millstein

WHAT IS IN THE BOOK?

INTRODUCTION

CONVOLUTIONAL NEURAL NETWORK ARCHITECTURE

HOW CONVOLUTIONAL NEURAL NETWORKS WORK?

CONVOLUTIONAL NEURAL NETWORKS APPLICATIONS

CONVOLUTION OPERATOR

CHAPTER 1: CONVOLUTIONAL NEURAL NETWORK LAYERS

CONVOLUTIONAL LAYER

SPATIAL ARRANGEMENT

STRIDE AND PADDING

PARAMETER SHARING

MATRIX MULTIPLICATION

POOLING LAYERS

FULLY-CONNECTED LAYER

NON-LINEARITY RELU

CHAPTER 2: TRAINING CONVOLUTIONAL NEURAL NETWORKS USING BACKPROPAGATION

APPLYING DROPOUT

THE OVERALL TRAINING PROCESS

CHAPTER 3: BUILDING A CONVOLUTIONAL NEURAL NETWORK

CREATING LAYERS

MAKING PREDICTIONS

CALCULATING LOSS FUNCTIONS

CONFIGURING MODELS

ADDING EVALUATION METRICS

TRAINING AND EVALUATING MNIST CLASSIFIERS

CREATING ESTIMATORS

TRAINING, EVALUATING AND RUNNING MODELS

CHAPTER 4: IMAGE CLASSIFICATION WITH CONVOLUTIONAL NEURAL NETWORKS

MAKING PREDICTONS FROM TRAINED MODELS

LAST WORDS

Copyright © 2018 by Frank Millstein- All rights reserved.

This document is geared towards providing exact and reliable information in regards to the topic and issue covered. The publication is sold with the idea that the publisher is not required to render accounting, officially permitted, or otherwise, qualified services. If advice is necessary, legal or professional, a practiced individual in the profession should be ordered.

From a Declaration of Principles which was accepted and approved equally by a Committee of the American Bar Association and a Committee of Publishers and Associations.

In no way is it legal to reproduce, duplicate, or transmit any part of this document by either electronic means or in printed format. Recording of this publication is strictly prohibited, and any storage of this document is not allowed unless with written permission from the publisher. All rights reserved.

The information provided herein is stated to be truthful and consistent, in that any liability, in terms of inattention or otherwise, by any usage or abuse of any policies, processes, or directions contained within is the solitary and utter responsibility of the recipient reader. Under no circumstances will any legal responsibility or blame be held against the publisher for any reparation, damages, or monetary loss due to the information herein, either directly or indirectly.

Respective authors own all copyrights not held by the publisher.

The information herein is offered for informational purposes solely and is universal as so. The presentation of the information is without contract or any type of guarantee assurance.

The trademarks that are used are without any consent, and the publication of the trademark is without permission or backing by the trademark owner. All trademarks and brands within this book are for clarifying purposes only and are owned by the owners themselves, not affiliated with this document.

INTRODUCTION

You probably already know that deep learning and machine learning are a broad computer science branch, which studies the design and architecture of models that can learn.  Deep learning is a broad subfield of MI or machine learning which is entirely inspired by neural networks that are inspired by human or biological neural networks.

One specific type of artificial neural network is the CNN or convolutional neural network, that is commonly referred to as ConvNet. It is a type of a feed-forwards deep neural network. You probably already know that feed-forward neural networks are the quintessential deep learning algorithms referred to as MLPs or multi-perceptron. They are called feed-forwards as information contained in them flows right through the models. They are called feed-forward as there are no feedback connections and the outputs of the algorithm are fed back into itself.

Convolutional neural networks are entirely inspired by the human or biological visual cortex. The cortex contains very small regions of cells, which are highly sensitive to certain areas of the human visual field. This main idea of CNNs was to expand upon an experiment that was performed by Hubel and Wiesel back in 1962. In this experiment, Hubel and Wiesel showed that specific brain neurons can be fired or activated when there is the presence of edges of very specific orientation such as horizontal or vertical edges.

For instance, certain neurons when exposed to vertical sides times show a horizontal edge. The two scientists, Hubel and Wiesel, later found that these neurons were, in fact, well ordered in columnar fashion, so together they were able to produce correct visual perception. The Hubel and Wiesel idea of a specialized component contained inside a system that has specific tasks, is the main concept used in convolutional neural networks.

When it comes to the broad field of computer vision, CNNs have been one of the most important innovations. Convolutional neural networks have performed much better than those traditional computer vision techniques and tools. CNNs have also proven to be amazingly successfully in many other computer vision areas like object detection, image classification, segmentation, and face recognition. CNNs are doing great in