In my undergraduate dissertation back in 2016 I studied how viable it would be to take a state-of-the-art neural network that had previously only existed in academic research and implement it into an open source machine learning library to use for a real world purposes and see if it retained performance.

I had never done anything with machine learning before and I have no background in mathematics. By summer 2017 I had read dozens of books, research papers and watched dozens of videos and I had finished with 13,000 words on the subject and a distinction grade.

This research forms the foundation of the book I am writing – Practical Deep Learning Fundamentals.

Approachable and highly practical

My aim is to create an approachable and highly practical guide to machine learning, with emphasis on the cutting edge techniques utilising neural networks, in a way which is accessible to readers with only a high school level background in math. My ultimate aim is to reduce the learning curve associated with machine learning.

You don’t need to be “good at math” to understand neural networks or machine learning to implement working models in open source libraries. I’m proof.

Contents

The book will guide readers through introductory fundamentals and use practical tutorials to build understanding, leading to three project chapters solving real world problems with neural networks in open source libraries. The final chapters will provide information for the reader to continue learning with other resources and a chapter as a primer to the math of neural networks, so the reader can have a greater understanding should they wish to delve deeper and read academic research.

I also intend to include case studies on how different machine learning techniques, including neural networks, are being used by businesses and academics. This might be useful for readers looking for inspiration, or general interest.

With the demand for machine learning practitioners growing I believe this book is of value to those looking to change their career direction or break into machine learning, as well as those looking to implement machine learning into personal projects or just to learn about machine learning.

Draft Contents

Chapter One: Preamble

Chapter Two: Introduction to Machine Learning

Chapter Three: Exploring Machine Learning Libraries and Methods

Chapter Four: Artificial Neural Networks

Chapter Five: Training Artificial Neural Networks

Chapter Six: Deep Learning

Chapter Seven: Evaluating and Tuning Neural Networks

Chapter Eight: Building Datasets

Chapter Nine: Project One: Sentiment Analysis in TensorFlow

Chapter Ten: Project Two: Computer Vision

Chapter Eleven: Project Three: Reinforcement Learning

Chapter Twelve: The State of Machine Learning

Chapter Thirteen: A Primer on Neural Network Math

This is a draft of the contents at the time of posting. There will likely be changes in the future, titles may change to become more descriptive.