Alexey Grigorev - Machine Learning Bookcamp: Build a portfolio of real-life projects

Build a portfolio of real-life projects

Alexey Grigorev

Foreword by Luca Massaron

To comment go to liveBook

Manning

Shelter Island

For more information on this and other Manning titles go to

www.manning.com

For online information and ordering of these and other Manning books, please visit www.manning.com. The publisher offers discounts on these books when ordered in quantity.

For more information, please contact

Special Sales Department

Manning Publications Co.

20 Baldwin Road

PO Box 761

Shelter Island, NY 11964

Email: orders@manning.com

2021 by Manning Publications Co. All rights reserved.

No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by means electronic, mechanical, photocopying, or otherwise, without prior written permission of the publisher.

Many of the designations used by manufacturers and sellers to distinguish their products are claimed as trademarks. Where those designations appear in the book, and Manning Publications was aware of a trademark claim, the designations have been printed in initial caps or all caps.

Recognizing the importance of preserving what has been written, it is Mannings policy to have the books we publish printed on acid-free paper, and we exert our best efforts to that end. Recognizing also our responsibility to conserve the resources of our planet, Manning books are printed on paper that is at least 15 percent recycled and processed without the use of elemental chlorine.

ISBN: 9781617296819

Ive known Alexey for more than six years. We almost worked together at the same data science team in a tech company in Berlin: Alexey started a few months after I left. Despite that, we still managed to get to know each other through Kaggle, the data science competition platform, and a common friend. We participated on the same team in a Kaggle competition on natural language processing, an interesting project that required carefully using pretrained word embeddings and cleverly mixing them. At the same time, Alexey was writing a book, and he asked me to be a technical reviewer. The book was about Java and data science, and, while reading it, I was particularly impressed by how carefully Alexey planned and orchestrated interesting examples. This led soon to a new collaboration: we coauthored a project-based book about TensorFlow, working on different projects from reinforcement learning to recommender systems that aimed to be an inspiration and example for the readers.

When working with Alexey, I noticed that he prefers to learn things by doing and by coding, like many others who transitioned to data science from software engineering.

Therefore, I wasnt very surprised when I heard that he had started another project-based book. Invited to provide feedback on Alexeys work, I read the book from its early stages and found the reading fascinating. This book is a practical introduction to machine learning with a focus on hands-on experience. Its written for people with the same background that Alexey hasfor developers interested in data science and needing to quickly build up useful and reusable experience with data and data problems.

As an author of more than a dozen books on data science and AI, I know there are already a lot of books and courses on this topic. However, this book is quite different. In Machine Learning Bookcamp, you wont find the same dj vu data problems that other books offer. It doesnt have the same pedantic, repetitive flow of topics, like a route already traced on maps that always leads to places that you already know and have seen.

Everything in the book revolves around practical and nearly real-world examples. You will learn how to predict the price of a car, determine whether or not a customer is going to churn, and assess the risk of not repaying a loan. After that, you will classify clothing photos into T-shirts, dresses, pants, and other categories. This project is especially curious and interesting because Alexey personally curated this dataset, and you can enrich it with the clothes from your own wardrobe.

By reading this book, of course, you are expected to apply machine learning to solve common problems, and you will use the simplest and most efficient solutions to achieve the best results. The first chapters begin by examining basic algorithms such as linear regression and logistic regression. The reader then gradually moves to gradient boosting and neural networks. Nevertheless, the strong point of the book is that, while teaching machine learning through practice, it also prepares you for the real world. You will deal with unbalanced classes and long-tail distributions, and discover how to handle dirty data. You will evaluate your models and deploy them with AWS Lambda and Kubernetes. And these are just a few of the new techniques you learn by working through the pages.

Thinking with the mind-set of an engineer, you can say that this book is arranged so that youll get the core 20% knowledge that covers 80% of being a great data scientist. More importantly, Ill add that youll be also reading and practicing under Alexeys guidance, which is distilled by his work and Kaggle experience. Given such premises, I wish you a great journey through the pages and the projects of this book. I am sure that it will help you find the best way to approach data science and its problems, tools, and solutions.

Luca Massaron

I started my career working as a Java developer. Around 20122013, I became interested in data science and machine learning. First, I watched online courses, and then I enrolled in a masters program and spent two years studying different aspects of business intelligence and data science. Eventually, I graduated in 2015, and started working as a data scientist.

At work, my colleague showed me Kagglea platform for data science competitions. I thought, With all the skills I got from courses and my masters degree, Ill be able to win any competition easily. But when I tried competing, I failed miserably. All the theoretical knowledge I had was useless on Kaggle. My models were awful, and I ended up on the bottom of the leaderboard.