Clustering Analysis in SPSS

1
Clustering Analysis in SPSS

• Multivariate Data Analysis Prediction for
  identifying groups
• Factor Analysis
• TwoStep Cluster Analysis
• K-means Cluster Analysis
• Hierarchical Cluster Analysis
• Discriminant Analysis

2
Hierarchical Cluster Analysis

• - Take clusters from a single record and form
  groups until all clusters are merged.
• - over 40 measures of similarity or
  dissimilarity,
• - standardize data using several methods and
  cluster cases or variables. - generate distance
  or similarity measures using the proximities
  procedure.
• - display statistics at each stage to select the
  best solution for datasets that are smaller in
  number.
• Example A market researcher could use
  Hierarchical Cluster Analysis to identify types
  of television shows that attract similar
  audiences for each show type. The organisation
  could cluster TV shows into homogenous groups
  based on viewer characteristics to identify
  segments for advertising.

3
K-means Cluster Analysis

• - group data from larger datasets, such as
  customer mailing lists.
• - assumes data fall into a known number of
  clusters. Given this number, the procedure will
  assign cases to clusters.
• - either update cluster centers iteratively
• - or classify only.
• - Save cluster memberships, distance information
  and final cluster centers.
• Example A market researcher might want to
  cluster cities into homogeneous groups using
  K-means Cluster Analysis to find comparable
  cities to test marketing strategies.

4
TwoStep Cluster analysis

• - work with very large datasets
• - can handle both continuous and categorical
  variables
• - Steps
• 1. pre-cluster the records into many small
  sub-clusters.
• 2. cluster the sub-clusters created in the
  pre-cluster step into the desired number of
  clusters.
• - If the desired number of clusters is unknown,
  
• it automatically finds the proper number of
  clusters.
• Example can be applied to data that describe
  customer buying habits, gender, age, income, etc.
  Then, tailor your marketing and product
  development strategy to each consumer group to
  increase sales and build brand loyalty.