Week 1

Question 1

difference between the Statisticians and Machine learning

Answer 1

Statisticians tend to start by making modelling assumptions about how the data is generated. Generally
these assumptions then give a mathematical framework in which to answer specific questions.

Machine learning people tend to treat the mechanism that generates the data as unknown (or
unknowable) and are happy to use any algorithmic model that gets the job done

Question 2

Key steps in data mining

Answer 2

1. Collect data (or get given it).
2. Wrangle the data into shape.
3. Train models (the more the better!)
4. Choose the best model.
5. Use the best model for prediction

Question 3

Data wrangling

Answer 3

data wrangling consists of doing everything necessary to get datasets ‘tidy’ and ready for
modelling.

Question 4

choose variables (columns) by name

Answer 4

select()

Question 5

to choose observations (rows) by value.

Answer 5

filter()

Question 6

to add new variables based of existing variables

Answer 6

mutate()

Question 7

to reduce multiple values down to a single summary.

Answer 7

summaries()

Question 8

changes the order of rows

Answer 8

arrange()

Question 9

If you want to rename a column while keeping the other columns

Answer 9

rename()

Question 10

You can remove grouping

Answer 10

ungroup()

Question 11

The function adds a count column instead of summarising

Answer 11

add_count()

Question 12

the function useful for finding the top (or bottom) few entries.

Answer 12

slice_min()
and
slice_max()

Question 13

function can be used to take a random sample

Answer 13

slice_sample()