Читать книгу AWS Certified Machine Learning Study Guide - Shreyas Subramanian - Страница 17

Anybody who wants to pass the AWS Certified Machine Learning Specialty exam may benefit from this book. This book is also helpful for business and IT professionals who want to learn how ML is practically used in the industry and pivot their careers toward an ML-centric role such as a data scientist or ML engineer working on AWS. We include a number of practical case studies, industry best practices, and architecture patterns that we have seen used in industry today from our engagements with hundreds of AWS customers. This book is also essential for data scientists, engineers, and other data professionals who are curious about how you can build, train, and deploy models at scale on AWS.

This book assumes some familiarity with ML and with AWS. If you are completely new to machine learning, we recommend that you first learn some basic ML concepts since this book is mainly focused on the practical aspects of building ML solutions. There are several great resources that cover ML foundations, particularly for building statistical models and for deep learning. Two of our favorites are Aurélion Géron's Hands-on Machine Learning with Scikit-learn and TensorFlow (O'Reilly Publishing) and Francois Chollet's Deep Learning with Python (Manning, 2017). There are also several awesome blogs on Medium.com and TowardsDataScience.com. Finally, we also recommend a number of industry blogs from leading tech companies like Uber, Google, Facebook, Amazon, Airbnb, and others on how they deploy large-scale ML solutions to have a holistic understanding of the industry landscape in this space.

As a practical matter, you'll need a laptop or desktop with which to practice and learn in a hands-on way. This book does not cover labs, and there is no substitute for hands-on experience. Go get familiar with AWS ML services such as SageMaker, as well as the AI services, before taking the test. We also recommend that you explore some public datasets, engineer features, and train simple models as well as some deep learning models.