IFN580 Week 2: Data to Features (7%)

Question 1

What are the two main types of variables in machine learning?

Answer 1

Categorical and numerical

Question 2

What are the subtypes of categorical variables?

Answer 2

Binary, nominal, ordinal

Question 3

What are the subtypes of numerical variables?

Answer 3

Discrete, continuous

Question 4

What is a binary variable?

Answer 4

Two categories; yes/no, 0/1, true/false

Question 5

What is a nominal variable?

Answer 5

Unordered categories: states, names of things (e.g., colour)

Question 6

What is an ordinal variable?

Answer 6

Ordered categories: low/medium/high, small/medium/large

Question 7

What is a discrete variable?

Answer 7

Variables that are whole numbers, countable (e.g., test score, postal code)

Question 8

What is a continuous variable?

Answer 8

Variables that are “whole numbers” (e.g., temperature, weight, height)

Question 9

What is data pre-processing?

Answer 9

Techniques used to clean data, in order to improve its quality

Question 10

What is noise in data?

Answer 10

Incorrect data, indicated by outliers

Question 11

What is missing data?

Answer 11

Data that is not present or that has been deleted during noise detection

Question 12

What methods can be used to correct inconsistent data?

Answer 12

Binning, clustering, rregression

Question 13

What is feature engineering?

Answer 13

The process of creating or modifying features to improve model performance

Question 14

Some examples of feature engineering include?

Answer 14

One-hot encoding, scaling