CLUSTER ANALYSIS AND MEDICI EFFECT

GROUP F   CLUSTER ANALYSIS - session 4

-By Amrita Bazray (14005)


In this class we studied about cluster analysis, its purpose and how to use it in the business context effectively through innovative thinking as mangers with a given set of information.So what is it?Firstly a cluster is a group of relatively homogeneous cases or observations.Cluster analysis is a statistical method of partitioning a sample into homogeneous classes to produce an operational classification which identifies groups of samples that behave similarly or show similar characteristics into meaningful taxonomies, groups, or clusters. It is a data reduction tool that creates subgroups that are more manageable than individual datum.
It makes no distinction between dependent and independent variables.·         Hierarchal- for more than 50 objects
·         Divisive clustering/ top-down clustering
·         Agglomerative clustering- successively merge (or agglomerate) pairs of clusters until all clusters have been merged into a single cluster that contains all documents. Also called HAC.
·         K-means-  for less than 50 objects

Procedure:

1.      Selection of variables- Based on the objective of clustering.
2.      Distance measurement- Difference between the two individual objects. Data can be Interval, binary and counts. (Euclidean distance, Block etc.). For counts we use Chi square. For Binary we use squared Euclidean distance method. For business purposes usually we have data in the form of either interval or binary.
3.      Clustering criteria- Methods used are nearest and furthest neighbor method and centroid clustering (mostly used)
If we discuss the first method of cluster analysis i.e. Hierarchical cluster analysis, we follow the above given procedure. We were taught that to carry out a hierarchical cluster analysis we usually use Ward’s method applying Euclidean Distance as the distance or similarity measure.

The results of the cluster analysis are shown by a dendrogram, which lists all of the samples and it indicates at what level of similarity any two clusters were joined.

Medici effect- author Frans Johansson. It relates to the concept of innovation and perspective. In this book he shares many interesting stories of cross-pollination between disciplines that result in explosion of new ideas and innovation. It is recommended that all students should read this book.