Machine Learning for the Quantified Self - Summary Slides

Machine Learning for the Quantified Self - Slides Summary


Lecture 1: Introduction and Basics of Sensory Data
Quantified Self definition
●​ Term first coined by Gary Wolf and Kevin Kelly in Wired Magazine
●​ Swan (2013): “The quantified self is any individual engaged in the self-tracking of any
kind of biological, physical, behavioral, or environmental information. There is a proactive
stance toward obtaining information and acting on it.”
●​ We: “The quantified self is any individual engaged in the self-tracking of any kind of
biological, physical, behavioral, or environmental information. The self-tracking is driven
by a certain goal of the individual with a desire to act upon the collected information.”

Quantified Self: measurements
●​ Augemberg (2012):




Quantified Self: why?
●​ Choe, 2014:
○​ Interview with 52 quantified selves
○​ Three categories:
■​ Improved health (cure or manage a condition, execute a treatment plan,
achieve a goal)
■​ Improve other aspects of life (maximize work performance, be mindful)
■​ Find new life experiences (have fun, learn new things)
●​ Gimpel, 2013:
○​ Identify “Five-Factor Framework of Self-Tracking Motivations”:
■​ Self-healing (become healthy)
■​ Self-discipline (rewarding aspects of it)
■​ Self-design (control and optimize “yourself”)
■​ Self-association (associated with movement)
■​ Self-entertainment (entertainment value)




1

,Machine Learning for the Quantified Self - Slides Summary


Quantified Self: Arnold and Bruce
●​ Use two running examples
○​ Arnold:
■​ Loves sports
■​ Wants to participate in IRONMAN
■​ Gadget freak
■​ Smart phone/watch/...
■​ Electronic scale
■​ Chest strap
■​ …...
○​ Bruce:
■​ Diabetic
■​ Susceptible for depression
■​ Smart watch
■​ Device to measure blood glucose level
■​ ......

Moving on the machine learning
●​ Machine learning: “Machine learning is to automatically identify patterns from data”
●​ What could we learn for Arnold and Bruce?
○​ Arnold:
■​ Advising the training to make most progress towards a certain goal based
on past outcomes of training.
■​ Forecasting when a certain running distance will be feasible based on the
progress made so far and the training schedule.
○​ Bruce:
■​ Predict the next blood glucose level based on past measurements and
activity levels.
■​ Determine when and how to intervene when the mood is going down to
avoid a spell of depression.
■​ Finding clusters of locations that appear to elevate one’s mood.

Why is the Quantified Self so different?
●​ Sensory noise
●​ Missing measurements
●​ Temporal data
●​ Interaction with a user
●​ Learn over multiple datasets




2

,Machine Learning for the Quantified Self - Slides Summary


Basic Terminology
●​ A measurement is one value for an attribute recorded at a specific time point.




●​ A time series is a series of measurements in temporal order.




●​ Machine learning terminology is assumed to be known, for your convenience:
○​ Supervised learning is the machine learning task of inferring a function from a
set of labeled training data
○​ In unsupervised learning, there is no target measure (or label), and the goal is
to describe the associations and patterns among the attributes
○​ Reinforcement learning tries to find optimal actions in a given situation so as to
maximize a numerical reward that does not immediately come with the action but
later in time

Mathematical notation




3

, Machine Learning for the Quantified Self - Slides Summary


Overview of the course




Dataset
●​ During the course we will use a running example provided by CrowdSignals.io
●​ People share their mobile sensor data (smart phone and smart watch) and get paid for
annotating their data with activities




Mobile phone measurements (examples)
●​ Accelerometer
○​ Measures the changes in forces upon the phone in the x-y-z plane
●​ Gyroscope
○​ Orientation of the phone compared to the earth’s surface
●​ Magnetometer
○​ Measures x-y-z orientation compared to the earth’s magnetic field




4