Cook D.J., Krishnan N.C. Activity Learning: Discovering, Recognizing, and Predicting Human Behavior from Sensor Data

Wiley, 2015. — 288 p. — ISBN: 9781118893760The goal of this book is to define the notion of an activity model learned from sensor data and to present the key algorithms that form the core of the field. While many current efforts have contributed to the ability to automatically recognize known activities, there are other aspects of activity learning that we want to include in this treatment. These include discovering activity patterns from unlabeled data and pre- dicting when specific activities will occur in the future, as well as mapping sensor event sequences to predefined activity labels. We also want to discuss the challenges that are faced when these theoretical techniques are applied to real-world problems and suggest methods for addressing the challenges.
This book is designed for an interdisciplinary audience who would like to use or design activity learning techniques. As a result, most of the ideas are presented from the ground up, with little assumptions about the background of the reader. Ideally, the book will provide some helpful background and guidance to researchers, under- graduate or graduate students, or practitioners who want to incorporate the ideas into their own work.