Book - Chapter 7 Analytical Theory Classification

Question 1

What applications does classification appear in

Answer 1

Data mining

Question 2

What is the primary task of a classifier

Answer 2

To assign class labels to new observations

Question 3

Are classification method supervised or unsupervised

Answer 3

Supervised

Question 4

What is another name for a decision tree

Answer 4

Prediction tree

Question 5

What is the input variable of a decision tree

Answer 5

Categorical or continuous

Question 6

In a decision tree structure what is a test point

Answer 6

A node

Question 7

What is a node without further branches called

Answer 7

A leaf node

Question 8

What do leaf nodes return

Answer 8

They return class labels and, in some implementations, they return the probability scores

Question 9

What are the two varieties of decision trees

Answer 9

Classification trees and regression trees

Question 10

What are classification trees

Answer 10

They usually apply to output variables that are categorical for example often binary yes or no

Question 11

What are regression trees

Answer 11

They can apply to output variables that are numerical continuous, such as the predicted price of a consumer good or the likely heard a subscription will be purchased

Question 12

What does the term branch mean in decision trees

Answer 12

Refers to the outcome of a decision and is visualised as a line connecting two Nodes

Question 13

What happens if the decision is numerical

Answer 13

The greater than branch is usually placed on the right

Question 14

What is an internal node

Answer 14

Are the dissertation or test points. Each internal note refers to an input variable or an attribute

Question 15

What is the top internal node called

Answer 15

The root

Question 16

What is the depth of a node

Answer 16

Is the minimum number of steps required to reach the node from the root

Question 17

What are short trees also known as

Answer 17

Weak learners or base learners

Question 18

What’s on in ensemble Mefford

Answer 18

They use multiple predictive models to vote, and decisions can be made based on the combination of the votes

Question 19

Gave examples of ensemble methods

Answer 19

Random forest, bagging, and boasting

Question 20

What is the simplest short tree called

Answer 20

Decision stump

Question 21

At each split what does the decision tree algorithm do

Answer 21

It picks the most informative attribute out of the remaining attributes

Question 22

How is the most informative attribute determined

Answer 22

By measures such as entropy and information gain

Question 23

What does entropy measure

Answer 23

The impurity of an attribute

Question 24

What does information gain measure

Answer 24

The purity of an attribute

Question 25

When do you achieve maximum entropy

Answer 25

When all class labels are equally probable

Question 26

What is conditional entropy always

Answer 26

Less than or equal to the base Entropy

Question 27

What is information gain defined as

Answer 27

The difference between base Entropy and conditional entropy

Question 28

What is Bayes theorem

Answer 28

Gives a relationship between the probabilities of two events and their conditional probabilities

Question 29

What is a naive Bayes classifier

Answer 29

Assumes that the presence or absence of a particular feature of a class is unrelated to the presence or absence of other features

Question 30

What are the input variables of naive Bayes

Answer 30

Categorical and I’ll discreet

Question 31

What is the output of naive Bayes

Answer 31

Class label and its corresponding probability score. The probability score is not the true probability of the class label, but it’s proportional to the true probability

Question 32

What is naive Bayes most commonly used for

Answer 32

Spam filtering

Question 33

What is Bayes theorem

Answer 33

The conditional probability of event C occurring, given that event A has already occurred, is to noted as P (C|A)

Question 34

What should a good classifier have

Answer 34

A large true positive and true negative and a small (ideally zero) numbers for false positives and false negatives

Question 35

What does accuracy mean

Answer 35

Defining the rate at which a model has classified the records correctly

Question 36

What is recall

Answer 36

The percentage of positive instances that were correctly identified