Book Image

Journey to Become a Google Cloud Machine Learning Engineer

By : Dr. Logan Song
Book Image

Journey to Become a Google Cloud Machine Learning Engineer

By: Dr. Logan Song

Overview of this book

This book aims to provide a study guide to learn and master machine learning in Google Cloud: to build a broad and strong knowledge base, train hands-on skills, and get certified as a Google Cloud Machine Learning Engineer. The book is for someone who has the basic Google Cloud Platform (GCP) knowledge and skills, and basic Python programming skills, and wants to learn machine learning in GCP to take their next step toward becoming a Google Cloud Certified Machine Learning professional. The book starts by laying the foundations of Google Cloud Platform and Python programming, followed the by building blocks of machine learning, then focusing on machine learning in Google Cloud, and finally ends the studying for the Google Cloud Machine Learning certification by integrating all the knowledge and skills together. The book is based on the graduate courses the author has been teaching at the University of Texas at Dallas. When going through the chapters, the reader is expected to study the concepts, complete the exercises, understand and practice the labs in the appendices, and study each exam question thoroughly. Then, at the end of the learning journey, you can expect to harvest the knowledge, skills, and a certificate.
Table of Contents (23 chapters)
1
Part 1: Starting with GCP and Python
4
Part 2: Introducing Machine Learning
8
Part 3: Mastering ML in GCP
13
Part 4: Accomplishing GCP ML Certification
15
Part 5: Appendices
Appendix 2: Practicing Using the Python Data Libraries

Introduction to TensorFlow

TensorFlow is an end-to-end open source platform for ML, developed by Google Brain, and it is one of the most widely used ML frameworks by data scientists.

TensorFlow flow tensors – TensorFlow’s name is directly derived from its core framework components: tensors. Let’s start by understanding tensors.

Understanding the concept of tensors

A tensor is a container that holds data of various sizes and shapes in an N-dimensional space. A tensor can be originated from the input data or a computation of the input data. In ML, we call the tensor components features. A tensor has three main characters to describe itself, called a tensor’s rank, shape, and dtype as follows:

  • Rank is the number of directions
  • Shape is the number of elements in each direction
  • Dtype is the data type

The rank of a tensor specifies the number of directions being measured for a tensor. From the number of ranks, a tensor can be categorized...