Funk Burkhardt - Machine Learning for the Quantified Self: on the art of learning from sensory data

Springer International Publishing AG 2018

Mark Hoogendoorn 1 and Burkhardt Funk 2

Before diving into the terminology and defining the core concepts used throughout this book, let us first start with two fictive, yet illustrative, examples that we will return to regularly throughout this book.

The first example involves a person called Arnold . Arnold is 25 years old, loves to run and cycle, and is a regular visitor of the gym. His ultimate goal is to participate in an IRONMAN triathlon race consisting of 3.86 kilometers of swimming, 180 kilometers of cycling and running a marathon to wrap it all upa daunting task. Besides being a fan of sports, Arnold is also a gadget freak. This combination of two passions has resulted in what one could call an obsession to measure everything around his physical state. He always wears a smart watch to monitor his heart rate and activity level and carries his mobile phone during all of his activities, allowing for his position and movements to be logged continuously in addition to a number of other measurements. He also installed multiple training programs on his mobile phone to help him schedule workouts. On top of that he uses an electronic scale in his bathroom that logs his weight and a chest strap to measure his respiration during running and cycling. All of this data provides him with information about his current state which Arnold hopes can help him to reach his ultimate goal making it to the finish line during the Hawaiian IRONMAN championship.

Contrary to Arnold, whom you could call a measurement enthusiast, Bruce also measures a lot of things around his body, but for Bruce this out of necessity. Bruce is 45 years old and a diabetic. In addition, he regularly falls into a depression. Bruce previously had trouble regulating his blood glucose levels using the insulin injections he has to take along with each meal. Luckily for Bruce, new measurement devices support him in to tackle his problems. He has access to a fully connected blood glucose measurement device that provides him with advice on the insulin dose to inject. To work on his mental breakdowns, Bruce installed an app that regularly asks him to rate his mental state (e.g. how Bruce is feeling, what his mood is, how well he slept, etcetera). In addition, the app logs all of his activities supported by location tracking and activity logging on his mobile phone, as it is known that a lack of activity can lead to severe mental health problems. The app allows Bruce to pick up early signals on a pending mood swing and to make changes to avoid relapsing into a depression.

While Arnold and Bruce might be two rather extreme examples, they do illustrate the developments within the area of measurement devices: more and more devices are becoming available that measure an increasing part of our daily lives and well-being. Performing such measurements around ones self, quantifying ones current state, is referred to as the quantified self , which we will define more formally in the next section. This book aims to show how machine learning, also defined more precisely in this chapter, can be applied in a quantified self setting.