Scribd
Upload
10 views

Cluster Analysis

Uploaded by

Janani Ravinagarajan
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
10 views

Cluster Analysis

Uploaded by

Janani Ravinagarajan
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 1

Cluster Analysis

Cluster analysis is a grouping technique and useful tool for prediction. cluster analysis itself
is primarily a descriptive technique used for grouping similar data points, it can indirectly aid
in prediction through the following ways:
1. Target Group Identification
2. Feature Selection
3. Training Set Creation
4. Target variable estimation
5. Anomaly Estimation
Consider the below example. These are the different types of cars parked in a high-rise
apartment parking lot
Car Color Make State
TN 04 Z 3435 Red BMW Tamil Nadu
KA 19 P 8055 Teal Jaguar Karnataka
KL 1A P 1234 Aqua Mini Cooper Kerala
TS 06 Q 3211 Red Benz Telangana
AP 04 U 9789 Teal BMW Andhra Pradesh
22 BH 6517 A Red Benz Bihar
KA 04 Z 3435 Red BMW Karnataka
TN 19 P 8055 Teal Jaguar Tamil Nadu
TN 1A P 1234 Aqua Mini Cooper Tamil Nadu
AP 06 Q 3211 Red Benz Andhra Pradesh
AP 04 U 9009 Teal BMW Andhra Pradesh
22 BH 0007 A Aqua Jaguar Bihar

TN 04 Z 3435 STATE – TAMIL NADU


TS 06 Q 3211 TN 04 Z 3435
22 BH 6517 A TN 19 P 8055
KA 04 Z 3435 TN 1A P 1234
AP 06 Q 3211

MAKE – BMW So on and so forth


TN 04 Z 3435
AP 04 U 9789
KA 04 Z 3435
AP 04 U 9009

The variables can be formed in cluster formed. But there is no dependent variable, because of
which one can say that there is no objective function here. There is a choice to select which
ever attribute is to be recognized. Each group is homogeneous based on selected attribute
only.
K-Mean Clustering
It is prototype clustering. (Centroid and medoid) and K is the number of clusters. The data set
is clustered based on K-value.

You might also like