Inside Deep Learning_ Math, Algorithms, Models [MEAP]_Edward Raff.pdf

©Manning Publications Co. To comment go to liveBook MEAP Edition Manning Early Access Program Inside Deep Learning Math, Algorithms, Models Version 1 Copyright 2020 Manning Publications For more information on this and other Manning titles go to manning.com ©Manning Publications Co. To comment go to liveBook welcome I’m delighted that you’ve purchased the MEAP for my book Inside Deep Learning: Math, Algorithms, Models. This book is geared toward those with a firm programing background who have some machine learning (ML) experience but want to go deeper. You are in good shape to reach through this book if you are comfortable with the basics of calculus, linear algebra, and statistics that go into machine learning. If you can record the inputs and correct outputs for a task, deep learning can help you translate that task from human time and effort to an automated process. The input could be an image, and the output “cat”, or “dog” for example, would describe the images’ content. The input could be an English sentence, and the output a French sentence with the same meaning. That’s machine translation. The input could be the description “cat” and the output be an actual image of a cat! This input/output nature makes deep learning widely applicable to almost any domain you can think of. It's why I got into machine learning in the first place, giving me a chance to have an impact and contribute to solving real problems in almost any domain. Deep learning (also called neural networks) gives us tools to help improve the quality of life in big and small ways, and my sincere desire is to pass that along to you. By the end of this book, you should understand: • What deep learning is • How to build and modify deep learning models • Which “building blocks” you should look toward for a given problem To achieve this there are two overall parts to the book. In each chapter we will not only show how to implement the code for these techniques but