What you'll learn

How to use cluster analysis in data mining

About the various types of clusters

About the Marketing applications of cluster analysis

Implications of wide variety of clustering techniques

Use clustering in statistical analysis

Requirements

Basic knowledge of statistics is required. Some familiarity with data analysis will be considered as an added advantage though it is not a necessity.

Description

Cluster Analysis is a statistical tool which is used to classify objects into groups called clusters, where the objects belonging to one cluster are more similar to the other objects in that same cluster and the objects of other clusters are completely different. In simple words cluster analysis divides data into clusters that are meaningful and useful. Clustering is used mainly for two purposes – clustering for understanding and clustering for utility.Application of cluster analysisCluster analysis is used in many fields like machine learning, market research, pattern recognition, data analysis, information retrieval, image processing and data compression.Cluster analysis can help the marketers to find out distinct groups of their customer base.Cluster analysis is used in the field of biology to find out plant and animal taxonomies and categorize genes with similar characteristicsCluster analysis is used in an earth observation database to group the houses in a city according to the house type, value and location.Clustering can also be used to segment the documents on the web based on a specific criteriaIn data mining, cluster analysis is used to gain in-depth understanding about the characteristics of data in each cluster.Clustering MethodsClustering methods can be divided into the following categoriesPartitioning methodHierarchical MethodDensity based methodGrid Based MethodModel Based MethodConstraint Based MethodAdvantages of Cluster AnalysisGiven below are the advantages of cluster analysisCluster analysis gives a quick overview of dataIt can be used if there are many groups in dataCluster analysis can be used when there are unusual similarity measures to be doneCluster analysis can be added on ordination plots and it is good for the nearest neighboursApproaches to cluster analysisThere are a number of different approaches used to carry out cluster analysis which are divided into twoHierarchical Method – Agglomerative Methods and Divisive MethodsNon Hierarchical Method also known as K-means Clustering methodsCluster Analysis Course ObjectivesAt the end of this course you will be able to knowHow to use cluster analysis in data miningAbout the various types of clustersAbout the Marketing applications of cluster analysisImplications of wide variety of clustering techniquesUse clustering in statistical analysis