The book "Machine learning"

Hi, Habrozhiteli to us from the printing house finally came a novelty from Henrik Brink, Joseph Richards and Mark Feveravolph.

This book will allow programmers, data analysts, statisticians, data processing specialists and everyone else to apply machine learning to solve real-world problems, or at least just to understand what it is. Readers, without resorting to a deep theoretical study of specific algorithms, will gain practical experience in processing real data, modeling, optimization and scanning of machine learning systems. For those who are interested in the theory, we discuss the mathematical basis of machine learning, explain some algorithms and give links to materials for additional reading. The main focus is on practical results in solving problems.


The book is intended for those who want to apply machine learning to solve various problems. It describes and explains the processes, algorithms and tools related to the basic principles of machine learning. Attention is focused not on how to write popular algorithms, but on their practical application. Each stage of building and using machine learning models is illustrated by examples, the complexity of which varies from simple to medium level.
')

Book structure

Part I, “Sequence of actions in machine learning,” introduces the five stages of the basic sequence of machine learning:

• Chapter 1, “What is machine learning?”, Explains what machine learning is and why it is needed.

• Chapter 2, “Real Data,” discusses in detail the characteristic stages of data preparation for machine learning models.

• Chapter 3, Modeling and Prediction, teaches using common algorithms and libraries how to create simple ML models and generate forecasts.

• In Chapter 4, “Evaluating and Optimizing a Model,” ML models are discussed in detail in order to evaluate and optimize their performance.

• Chapter 5, “The Basics of Feature Design,” describes how to increase the amount of raw data using information from our task.

Part II “Practical Application” introduces techniques for scaling models, as well as techniques for extracting features from text, images, and time series, which increase the efficiency of solving many modern problems with machine learning. This part contains three chapters with practical examples.

• Chapter 6, “Example: A Tip for Taxi Drivers,” is the first one entirely devoted to the consideration of an example. We will try to predict the taxi driver's chances of getting a tip.

• Chapter 7, Advanced Feature Design, introduces the more sophisticated feature design techniques for extracting values ​​from text, images, and time series.

• In Chapter 8, An Example of Natural Language Processing, advanced feature design techniques are used to predict the tonality of film reviews.

• Chapter 9, “Scaling up the machine learning process,” introduces technicians that give ML systems the ability to work with large amounts of data, providing higher prediction speeds and reducing their waiting time.

• In Chapter 10, “A Digital Advertising Example,” a model is predicted on a large amount of data that predicts the probability of a transition by an advertising banner.

How to read this book

Those who do not yet have experience in the field of machine learning, chapters from the 1st to the 5th will be introduced to the processes of preparing and researching data, designing features, modeling and evaluating models. Python code examples use popular libraries such as pandas and scikit-learn. Chapters 6 through 10 include three practical examples of machine learning along with such advanced topics as feature design and optimization. Since the basic computational complexity is encapsulated in libraries, these code snippets are easy to adapt to your own ML applications.

About the authors

Henrik Brink is a data processing and analysis specialist and software developer with extensive hands-on experience in machine learning in both production and research.

Joseph Richards is a senior researcher in applied statistics and predictive analytics. Henrik and Joseph co-founded Wise.io, a company that develops machine learning solutions for industry.

Mark Feverolf is the founder and president of Numinary Data Science, a company specializing in data management and predictive analytics. He worked as a statistician and developer of analytical databases in the fields of social sciences, chemical engineering, information system performance, production planning, cable television, and Internet advertising applications.

»More information about the book can be found on the publisher's website.
» Table of Contents
» Excerpt

For Habrozhiteley 25% discount coupon - Machine Learning