This book provides practical guidelines on how to use machine learning algorithms to create predictive models. It is very common to learn about machine learning in tutorials using toy, or small, datasets, which is very practical for learning the basic concepts but not enough when trying to apply what you've learned to real problems.
This book covers the main steps to be followed in order to develop predictive models based on machine learning algorithms. Data collection, data treatment, univariate and multivariate analysis, and the application of the most common machine learning algorithms are some of the steps described in this book. This is a programming book, containing several lines of code, therefore you could replicate all the steps described in it.
This book shows how there are no unique modeling possibilities; different existing options in each modeling step are key to achieving accurate and useful models.
The applications included in this book have been based on the financial sector. This is mainly because I am familiar with the information and the problems, and because there is a huge amount of data to apply several techniques to in a way that is representative of problems that can be found in real life.
The theoretical framework of the book is based on explaining the financial crisis and its causes. Would we be able to predict the next financial crisis? If not, at least you will learn very useful techniques for squeezing your data.
