Module 1 - Data and Foundations

Question 1

The 2 types of data

Answer 1

Quant: temperature (continuum), integer variables (discrete), etc.

Qual: toggles that switch betw. controls (logical/state), categorical (ie. red buttons)

Question 2

“Event”?

Answer 2

Event = a specific outcome with a probability

Question 3

Axioms of probabiltiy

Answer 3

P(S) = 1 (probability of some outcome S in sample space happening is 1)

0 <= P(S) <= 1

If I have 2 MUTUALLY EXCLUSIVE events: P(E1 U E2) = P(E1) + P(E2)

Question 4

Random variables

Answer 4

Attach numerical labels to an outcome

Question 5

Example of random variable

Answer 5

If random variable value is ‘grey’ then number assigned is 4

T is the random variable for any temperature value on a continuum scale

Question 6

Cumulative probability function for discrete random variable

Answer 6

F_X(R) - P(x <= k) = SUM(xi*Pi) (xi=i) for all i = 0 –> k

Question 7

Cumulative probability function for continuous random variable

Answer 7

f_X(a) = P(a - st.dev < x < a + st.dev)

Question 8

Cumulative distribution for continuous random var

Answer 8

P(x < k) = INT(f_X(x) dx)

Question 9

Expected value for discrete cases

Answer 9

xi*P_X(xi) –> sum for all xi and P_X(xi)

Question 10

Expected value for continuous cases

Answer 10

E{X} = INT(x*f_X(x) dx)

Question 11

Linear operator conditions for expected value

Answer 11

Commutative: E{X1 + X2} = E{X1} + E{X2}

Mult. by scalar: E{kX} = kE{X}

Question 12

Human-centred insights

Answer 12

Ethnography

Question 13

________ ________ brings in human-centred insights, which is another kind of _____.

Answer 13

Design thinking

Data

Question 14

Anthropogenic

Answer 14

Made by humans

Question 15

Problems with surveillance tech to collect data from the public

Answer 15

• Social policies bring in questions of privacy
• Who owns the tech? Are they ethical?
• Misidentification, inappropriate conclusions drawn from data
• Gaps in monetary network (too expensive)

Question 16

We need a ___________ data set to feed a model so it is trained for the appropriate task

Answer 16

Representative

Question 17

In engineering, we use data to help develop solutions that lie at the intersection of…

Answer 17

Technical feasibility, social desirability, financial viability

Question 18

Sources of variability

Answer 18

Ambient fluctuations

Instrumentation and measurement fluctuations

(For chem processes) Models and system/process representation

Question 19

Ambient fluctuations

Answer 19

Disturbances within a process or physical system

Human interventions

External disturbances

Question 20

Instrumentation and measurement fluctuations

Answer 20

Electronic noise

Physical location of instrument

Question 21

Models and system/process representation

Answer 21

Assumption of well-mixed is spatially distributed

Associated with simplification in model form

Question 22

Active collection

Answer 22

Make a series of planned moves on process

Increases info content, guarantees “causality”/”cause and effect” relationships

Question 22

Passive collection

Answer 22

Record process values without actively intervening

“Historical” databases = passively collected data, ie. browser history

Question 23

Role of statistical methods

Answer 23

1. Decision-making under uncertainty
2. Categorizing and modelling variability (ie. Poisson distr w/ mean, var)
3. Basis for “variability accounting” –> how it propogates
4. Data “microscope”
5. Effective presentation of results in graph/quant forms