DATA MINING

( Instructions: This Assignment must be submitted on Blackboard via the allocated folder. Email submission will not be accepted. You are advised to make your work clear and well-presented , marks may be reduced for poor presentation . You MUST show all your work . Late submission will result in ZERO marks being awarded. Identical copy from students or other resources will result in ZERO marks for all involved students. ) ( Student Details: Name: ### CRN : ### ID: ### )

College of Computing and Informatics

a) What are outliers? List four applications of outlier detection.

b) What are the challenges of outlier detection?

a) How does PAM (K-medoids) form clusters; how does DBSCAN form clusters?

b) Assume you apply DBSCAN to the same dataset, but the examples in the dataset are sorted differently. Will DBSCAN always return the same clustering for different orderings of the same dataset? Give reasons for your answer.

( 1 Mark ) ( Learning Outcome(s): LO-2, 3 )Question Three

Measuring geodesic distance for the graph G in given figure, calculate the following:

i. Eccentricity

ii. Radius

iii. Diameter

iv. Peripheral vertex

( A D E C B F )

( 1 Mark ) ( Learning Outcome(s): LO-2 )Question Four

Why is it often necessary to do constraint-based clustering? Describe the terms hard constraint and soft constraint.

( Learning Outcome(s): LO-2, 3 )