Mastering Machine Learning with scikit-learn – Second Edition

Book Preface

In recent years, popular imagination has become fascinated by machine learning. The discipline has found a variety of applications. Some of these applications, such as spam filtering, are ubiquitous and have been rendered mundane by their successes. Many other applications have only recently been conceived, and hint at machine learning’s potential. In this book, we will examine several machine learning models and learning algorithms. We will discuss tasks that machine learning is commonly applied to, and we will learn to measure the performance of machine learning systems. We will work with a popular library
for the Python programming language called scikit-learn, which has assembled state-of-theart implementations of many machine learning algorithms under an intuitive and versatile API.

What this book covers

The Fundamentals of Machine Learning, defines machine learning as the study and design of programs that improve their performance of a task by learning from experience. This definition guides the other chapters; in each, we will examine a machine learning model, apply it to a task, and measure its performance.

Simple Linear Regression, discusses a model that relates a single feature to a continuous response variable. We will learn about cost functions and use the normal equation to optimize the model.

Classification and Regression with K-Nearest Neighbors, introduces a simple, nonlinear model for classification and regression tasks. Feature Extraction, describes methods for representing text, images, and categorical variables as features that can be used in machine learning models. Chapter 5, From Simple Linear Regression to Multiple Linear Regression, discusses a generalization of simple linear regression that regresses a continuous response variable onto multiple features.

From Linear Regression to Logistic Regression, further generalizes multiple linear regression and introduces a model for binary classification tasks. Naive Bayes, discusses Bayes’ theorem and the Naive Bayes family of classifiers, and compares generative and discriminative models. Nonlinear Classification and Regression with Decision Trees, introduces the decision tree, a simple, nonlinear model for classification and regression tasks. From Decision Trees to Random Forests and other Ensemble Methods, discusses three methods for combining models called bagging, boosting, and stacking.

The Perceptron, introduces a simple online model for binary classification. From the Perceptron to Support Vector Machines, discusses a powerful, discriminative model for classification and regression called the support vector machine, and a technique for efficiently projecting features to higher dimensional spaces. From the Perceptron to Artificial Neural Networks, introduces powerful nonlinear models for classification and regression built from graphs of artificial neurons. K-means, discusses an algorithm that can be used to find structures in unlabeled data.

Dimensionality Reduction with Principal Component Analysis, describes a method for reducing the dimensions of data that can mitigate the curse of dimensionality.

What you need for this book
The examples in this book require Python >= 2.7 or >= 3.3 and pip, the PyPA recommended tool for installing Python packages. The examples are intended to be executed in a Jupyter notebook or an IPython interpreter. ?????????, The Fundamentals of Machine Learning shows how to install scikit-learn 0.18.1, its dependencies, and other libraries on Ubuntu, Mac OS, and Windows.

Who this book is for
This book is intended for software engineers who want to understand how common machine learning algorithms work and develop an intuition for how to use them. It is also for data scientists who want to learn about the scikit-learn API. Familiarity with machine learning fundamentals and Python is helpful but not required.