Part I

Applied Math and Machine

Learning Basics

27

This part of the book introduces the basic mathematical concepts needed to

understand deep learning. We begin with general ideas from applied math that

enable us to deﬁne functions of many variables, ﬁnd the highest and lowest points

on these functions, and quantify degrees of belief.

Next, we describe the fundamental goals of machine learning. We describe how

to accomplish these goals by specifying a model that represents certain beliefs,

designing a cost function that measures how well those beliefs correspond with

reality, and using a training algorithm to minimize that cost function.

This elementary framework is the basis for a broad variety of machine learning

algorithms, including approaches to machine learning that are not deep. In the

subsequent parts of the book, we develop deep learning algorithms within this

framework.

28