lecture 1 - introduction

Question 1

definition: quantified self

1. swan
2. added

Answer 1

1. Swan: 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.
2. added: the self tracking is driven by a certain goal of the individual with a desire to act upon the collected information

• i.e., it is not about the data per-se, but about the person

Question 2

Gimpel (2013): five-factor framework of self-tracking motivations (FFFSM)

Answer 2

1. self-healing (become healthy)
2. self-discipline (rewarding aspects of it)
3. self-design (control and optimize ‘yourself’)
4. self-association (associated with movement)
5. self-entertainment (entertainment value)

Question 3

definition: machine learning

Answer 3

to automatically identify patterns from data

Question 4

why is the quantified self so different?

Answer 4

1. sensory data is noisy
2. a lot of missing measurements
3. the data is highly temporal in nature (and can even have multiple temporal sessions)
4. algorithms should enable the support and interaction with users without a long learning period
5. learning is over multiple temporal datasets (multiple quantified selves)

Question 5

definition: measurement

Answer 5

1. one value for an attribute recorded at a specific time point
2. series of measurements in temporal order

Question 6

definition: supervised learning

Answer 6

the machine learning task of inferring a function from a set of labeled training data

Question 7

definition: unsupervised learning

Answer 7

there is no target measure (or label), and the goal is to describe the associations and patterns among attributes

Question 8

definition: reinforcement learning

Answer 8

tries to find optimal actions in a given situation so as to maximise a numerical reward that does not immediately come with the action but later in time

Question 9

step size (Δt)

Answer 9

• represents one discrete time step

1. start at the earliest time point in the data
2. find all measurements for each single attribute associated with each interval [t, t+Δt)
3. we consider categorical features as a number of binary features (present or not)
4. combine the values (e.g., average)

Question 10

categorical values in step size

Answer 10

1. we can count whether at least one measurement of that value has been found in the time interval (binary)
2. or we can count the number of measurements that have been found for that value (sum)

Question 11

step size: longer engagement step

Answer 11

1. fewer missing values
2. less data, so less computationally expensive

Question 12

step size: shorter/fine grain data

Answer 12

1. more noise
2. more data

Question 13

choosing granularity has to do with

Answer 13

1. finding a balance between too many missing values vs too much noise
2. how precise are the measurements needed to make up a good prediction

Question 14

machine learning tasks

Answer 14

1. classification
2. regression

Question 15

Chloe (2014): categories that motivate conducting research with quantified selves

Answer 15

1. improved health: cure or manage a condition, execute a treatment plan, achieve a goal
2. improve other aspects of life: maximize work performance, be mindful
3. find new life experiences: have fun, learn new things