Personalized Machine Learning

Every day we interact with machine learning systems offering  individualized predictions for our entertainment, social connections,  purchases, or health. These involve several modalities of data, from  sequences of clicks to text, images, and social interactions. This book  introduces common principles and methods that underpin the design of  personalized predictive models for a variety of settings and modalities.  The book begins by revising 'traditional' machine learning models,  focusing on adapting them to settings involving user data, then presents  techniques based on advanced principles such as matrix factorization,  deep learning, and generative modeling, and concludes with a detailed  study of the consequences and risks of deploying personalized predictive  systems. A series of case studies in domains ranging from e-commerce to  health plus hands-on projects and code examples will give readers  understanding and experience with large-scale real-world datasets and  the ability to design models and systems for a wide range of  applications.