About Data Science

Question 1

Analytics

Answer 1

Based on exploring potential future events. You look for patterns

Question 2

Qualitative analytics

Answer 2

Intuition + analysis

Question 3

Quantitative analytics

Answer 3

Formulaes + algorithm

Question 4

Data Science

Answer 4

Valt buiten business analytics: Gaat om verbeteringen maar niet op basis van je bedrijfsdata

Question 5

Data Analytics

Answer 5

Valt buiten business analytics: Gaat om verbeteringen op basis van je data maar is buiten je bedrijfsdata

Question 6

Business Intelligence

Answer 6

Aims to explain past events using business data. It is a PRELIMINARY STEP of predictive analytics.

Question 7

First steps in data process

Answer 7

Data collection
Process that info
Pre-processing into workable format
Labeling (numerical, categorical)
Data cleansing

Question 8

Data shuffling

Answer 8

Prevents unwanted patterns by randomizing the data

Question 9

When do you know data is big data?

Answer 9

When looking at the graph you see that data is not stored in a way that is easy to oversee

When seeing a lot of lines intersect you are probably dealing with big data

Question 10

METRIC

Answer 10

MEASURE + BUSINESS MEANING

Question 11

KPI’s (Key Performance Indicators)

Answer 11

Metrics + business objectives

Question 12

Example Metric vs KPI:

Answer 12

Metric: Traffic of a page from your website that was visited by any type of user

KPI: Traffic generated only from users who have clicked on a link provided in you ad campaign

So KPI is more specific

Question 13

After business intelligence phase?

Answer 13

Predictive analytics –> Future outcomes

Question 14

Regression

Answer 14

A model used for quantifying causal relationships among the different variables included in you analysis

Question 15

Bow analogy (finding the best way to use the bow)

Model:

Data:

Target:

Optimization algorithm:

Answer 15

Model: The usage of the bow

Data: Quiver of arrows (they are all arrows but they differ a lot)

Target: Objective function (calculate how for from target)

Optimization algorithm: Mechanics that will improve the model’s performance (posture/how it holds the bow, etc.)

Question 16

Supervised learning

Answer 16

Training the algorithm resembles a teacher supervising her students

Question 17

Unsupervised learning

Answer 17

When not having the time or resouces

Question 18

Supervised and unsupervised data working together in bow example

Answer 18

Shoot a million arrows (which you don’t have time for to analyse)

Divide results into clusters

Use example of those clusters to determine the targets for those type of arrows

Run a supervised test to analyse the mechanics (previous bit)

Question 19

Example Qualitative vs. Quantitative analysis

Answer 19

Qualitative analysis: BI

Quantitative analysis: SWOT –> Because NOT data driven