Clustering

Which of the following are clustering algorithms?

• DBSCAN

• BIRCH

• OPTICS

• K-NN

How do you differentiate classification from clustering?

• Both are same

• Classification can be solved using supervised learning where as clustering is unsupervised learning

• Clustering can be solved using supervised learning where as classification is unsupervised learning

• Both unsupervised learning problems

• Both are supervised learning problems

Select all the correct statements below:

• DBSCAN is a hierarchical clustering algorithm

• Both DBSCAN and OPTICS are density based clustering algorithms

• BIRCH is a hierarchical clustering algorithm

• BIRCH, DBSCAN and OPTICS are density based clustering algorithms

• DBSCAN and OPTICS are hierarchical clustering algorithms

Select all the correct statements below:

• If e(eta)-neighborhood of a point p contains more than minimum points required, a new cluster with p as a core object is created

• Every object contained in a cluster is called a noise

• OPTICS is a hierarchical clustering alogrithm

• DBSCAN is a supervised learning algorithm

Which of the following algorithms computes a clustering order for cluster analysis?

• BIRCH

• DENCLUE

• OPTICS

• DBSCAN

Select the correct statement(s):

• A clustering order can be used to extract basic clustering information like cluster centers or arbitrary shaped clusters

• A clustering order provides intrinsic clustering structure

• DBSCAN algorithm uses the parameters named "Core Distance" and "Reach-ability Distance"

• OPTICS algorithm is a hierarchical clustering algorithm

Select the correct statement(s):

• If an object 'p' is a core object, the core distance of p is undefined.

• If an object 'p' is not a core object, the core distance of p is undefined.

• If p is not a core object, the reachability distance between p and q is undefined.

• If p is a core object, the reachability distance between p and q is undefined.

A clustering feature data structure in BIRCH consists of the following(s):

• Linear sum

• Square Sum

• number of data objects

• shape of the cluster

If A and B are two items in a dataset D. There are 1000 transactions in D and the item A is purchased 500 times. Items A and B are together purchased in 200 transactions, then support(A) is:

• 0.5

• 5/2

• 2/5

• 200/1000

If A and B are two items in a dataset D. There are 1000 transactions in D and the item A is purchased in all of these. Items A and B are together purchased in 200 transactions, then confidence(A=>B) is:

• 200

• .2

• 5

• .5