Deck 4: An Excel-Based Data Mining Tool

Question 1

Suppose that the predictiveness score for risk factor = medium risk is 0.50. How many domain instances have a value of medium risk for the risk factor attribute?

A) 10
B) 20
C) 30
D) 40

Answer 1

B

Question 2

The single best representative of a class.

A) mean
B) centroid
C) signature
D) prototype Answers to Chapter 4 Questions
Multiple Choice Questions

Answer 2

D

Question 3

A particular categorical attribute value has a predictiveness score of 0.3 and a predictability score of 0.3. The attribute value is

A) necessary but not sufficient for class membership.
B) sufficient but not necessary for class membership.
C) necessary and sufficient for class membership.
D) neither necessary nor sufficient for class membership.

Answer 3

D

Question 4

A particular categorical attribute value has a predictiveness score of 1.0 and a predictability score of 0.50. The attribute value is

A) necessary but not sufficient for class membership.
B) sufficient but not necessary for class membership.
C) necessary and sufficient for class membership.
D) neither necessary nor sufficient for class membership.

Question 5

A particular categorical attribute value has a predictiveness score of 0.5 and a predictability score of 1.0. The attribute value is

A) necessary but not sufficient for class membership.
B) sufficient but not necessary for class membership.
C) necessary and sufficient for class membership.
D) neither necessary nor sufficient for class membership.

Question 6

A dataset of 1000 instances contains one attribute specifying the color of an object. Suppose that 800 of the instances contain the value red for the color attribute. The remaining 200 instances hold green as the value of the color attribute. What is the domain predictability score for color = green?

A) 0.80
B) 0.20
C) 0.60
D) 0.40

Question 7

ESX represents the overall similarity of the exemplars contained in an individual class by computing a ____ score.

A) class resemblance
B) class predictability
C) class predictiveness
D) typicality

Question 8

A certain dataset contains two classes- class A and class B- each having 100 instances. RuleMaker generates several rules for each class.
One rule for class A is given as att1 = value1
# covered = 20
# remaining =60
What percent of the class A instances are covered by this rule?

A) 20
B) 40
C) 60
D) 70
E) 80

Question 9

The first row of an iDAV formatted file contains attribute names. The second row reflects attribute types. What is specified in the third row of an iDAV formatted file?

A) attribute predictability
B) attribute tolerance
C) attribute similarity
D) attribute usage

Question 10

This iDA component allows us to decide if we wish to process an entire dataset or to extract a representative subset of the data for mining.

A) preprocessor
B) heuristic agent
C) ESX
D) RuleMaker

Question 11

Which relationship is likely to be seen with an interesting clustering of data instances?

A) The domain resemblance score is greater than the resemblance scores for the individual clusters.
B) The domain resemblance score is equal to the average of the resemblance scores for the individual clusters.
C) The resemblance scores for all formed clusters are greater than zero.
D) The domain resemblance score is less than the resemblance scores for the individual clusters.

Question 12

What is the predictability score for the attribute value medium risk?

A) 0.10
B) 0.20
C) 0.25
D) 0.50