Scribd
Upload
47 views19 pages

SMA 3 Group 1 Cluster Analysis

Cluster analysis is used to classify objects or cases into relatively homogeneous groups. Objects in each cluster tend to be similar to each other and dissimilar to objects in the other clusters. The document discusses cluster analysis and provides examples of how it can be used in marketing and management questions, such as identifying customer usage levels or types of profitable products. It also compares cluster analysis to other analysis techniques such as discriminant analysis and factor analysis.

Uploaded by

Ngoc Tran Quang
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
47 views19 pages

SMA 3 Group 1 Cluster Analysis

Cluster analysis is used to classify objects or cases into relatively homogeneous groups. Objects in each cluster tend to be similar to each other and dissimilar to objects in the other clusters. The document discusses cluster analysis and provides examples of how it can be used in marketing and management questions, such as identifying customer usage levels or types of profitable products. It also compares cluster analysis to other analysis techniques such as discriminant analysis and factor analysis.

Uploaded by

Ngoc Tran Quang
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 19

CLUSTER

ANALYSIS
Understanding
the basic
concept of
cluster analysis
TODAY'S 1 Definition

AGENDA 2 Situation to be used

3 Management/ marketing questions

4 Cluster analysis and discriminant anaysis


Example:

I. DEFINITION A retail company may collect the following


information on households:
Household income
Household size
Cluster analysis: Head of household Occupation
Used to classify objects or cases into Distance from nearest urban area
relatively homogeneous groups. Objects in They can then feed these variables into a
each cluster tend to be similar to each clustering algorithm to perhaps identify the
other and dissimilar to objects in the other following clusters:
clusters. Cluster 1: Small family, high spenders
Cluster analysis is also called classification Cluster 2: Larger family, high spenders
analysis, or numerical taxonomy. Cluster 3: Small family, low spenders
Cluster 4: Large family, low spenders
2 CLUSTER
SITUATION:

Ideal Clustering Situation: A situation in which the


clusters are distinctly separated. Objects fall into
one cluster and there are no overlapping areas.

Practical Clustering Situation: A situation in which


the boundaries for some of the clusters are not
clear-cut, and the classification of some objects is
not obvious, because many of them could be
grouped into one cluster or another.
CLASSIFICATIONS
HIERARCHICAL NON-HIERARCHICAL OTHER
Agglomerative: Clusters The nonhierarchical Twostep
are formed by grouping clustering method is
objects into bigger and frequently referred to as k- Using for large data file, a
bigger clusters. means clustering. mixture of continuous and
categorical variables.
Divisive: Clusters are Using when knowing how
formed by dividing this many clusters we want and Can automatically
cluster into smaller and you have a moderately sized determine the optimal
smaller clusters. data set. number of clusters.

Using for small data set


and want to easily
examine solutions with
increasing numbers of
clusters.
Cluster analysis examines an entire set of
2. SITUATION TO
interdependent relationships and makes no
USE CLUSTER
ANALYSIS
distinction between variables.

Objects in a group are relatively similar in terms

of these variables.

The data can be vary.

Interval-Scaled variables: temperature, longitude

and latitude (clustering houses)


Binary variables: male and female
Nominal, Ordinal, and Ratio variables
Variables of mixed types

-> Cluster analysis is used during the exploratory research

phase when the researcher does not have any preformed

hypotheses.
3.
MANAGEMENT/ 1.How can a business identify the level of
MARKETING usage (from low to high) of customers ?

QUESTIONS 2.How can a business identify consumers


who use email in similar ways ?

3.How do I classify my products to


determine the appropriate service level?

Some relevant questions could be


answered:
Who is my organization’s best member?
Who is most likely to buy my services?
What types of products are most profitable?
CLUSTER

FACTOR ANALYSIS ANALYSIS

Explain the correlation with a data set

Address the heterogeneity in the

and understand how the variables

individual data sets.


relate to each other.

4. COMPARISION
4. COMPARISION

DISCRIMINANT
CLUSTER

ANALYSIS ANALYSIS
Prior knowledge of the group
There is no prior knowledge about the

membership for each object or case is


group membership for any objects.

required to develop classification


Groups/Clusters are suggested by the

rules. data.
SPSS
PRACTICES
Are You Ready?
1
As many groups as possible (no certain numbers of group required)
2
2 groups: label these groups and identify the number of students in each group.
2
2 groups: label these groups and identify the number of students in each group.

-There are 19 students classified in cluster 1 and 31


students classified in cluster 2.

-The final cluster centers table shows the mean of


each factor in two clusters.

-Label:
Cluster 1: Group of students whose decisions are
least influenced by factors.
Cluster 2: Group of students whose decisions are
most influenced by factors.
3
3 groups: label these groups and identify the number of students in each group.
3
3 groups: label these groups and identify the number of students in each group.

-There are 3 students classified in cluster 1, 27 students


classified in cluster 2, and 20 students classified in cluster 3.

-The final cluster centers table shows the mean of each factor
in three clusters.

-Label:
Cluster 1: Group of students whose decisions are least
influenced by factors.
Cluster 2: Group of students whose decisions are most
influenced by factors.
Cluster 3: Group of students whose decisions are medium
influenced by factors.
4
Optimal number of groups.
4
Optimal number of groups.
THANK
YOU!
Have a
great day
ahead.

You might also like