Data Analytics: Exploiting the Data Warehouse - Data Warehouse Systems: Design and Implementation - page 374

9.19 Comment on the dashboard design guidelines.
9.20 Define a dashboard using an example of an application domain that
you are familiar with.

9.7 Exercises

9.1 Consider the following training data about students:

StudID Age Country FamilyIncome Distance Finish
s1

1 local low 1 1

s2 0 local medium 0 0

s3 0 local high 1 0

s4 0 local medium 1 0

s5 4 foreigner medium 2 1

s6 3 foreigner medium 1 1

s7 3 foreigner low 1 2

s8 2 foreigner low 1 3

s9 1 local high 2 3

s10 0 local high 1 2

where the classes are as follows:

￿ Age indicates the age at which the student started the studies.
Possible values are as follows: 0 (between 17 and 21), 1 (between
22 and 26), 3 ( between 27 and 32), and 4 (older than 32).
￿ Country can have two values: local and foreigner .
￿ FamilyIncome can be low , medium ,and high .
￿ Distance indicates the distance that the student has to travel to go
to university. It can take values 0 (less than 1 mile), 1 (between 1
and 3 miles), and 2 (more than 3 miles).
￿ Finish indicates whether the student finished her studies in the years
planned for the corresponding career. It can take the values 0 (the
student finished her studies on time), 1 (the student finished at most
with 1-year delay), 2 (the student finished with 2 or more years of
delay), and 3 (the student abandoned her studies).

(a) Manually run the ID3 algorithm to build a decision tree over the
class Finish . Use the Gini index to partition the nodes.
(b) Use the K-means algorithm to generate three clusters of students.

9.2 Consider the Foodmart data warehouse of Fig. 6.4 :

(a) Build a decision tree model that predicts whether a new customer
is likely to order an item of a (sub)category X .Usethe Customer ,
Product , Product class ,and Sales tables in the Foodmart data