Cluster Analysis
Cluster analysis is a powerful data mining tool, especially in business as organizations can use it discrete customer groups, sales transactions, and other behaviors. For example, financial institutions can use cluster analysis to evaluate credit scores and fraudulent claims. The primary objective of cluster analysis is to identify similar groups, whereby similarity refers to a specific measure over collective characteristics.
Cluster analysis is mostly used in classification. Using cluster analysis in classification involves separating subjects into groups also knows as clusters so that each subject shares more similarities with other subjects in the group than the subjects outside the group. Cluster analysis is different from most statistical methods because it is used as an alternative in scenarios where there is no assumed principle or fact that researchers can use a research foundation (Setyaningsih, 2012).
Cluster analysis is performed during the explanatory phase of research because it cannot be used to distinguish between dependent and independent variables. On the contrary, this analysis technique is used to discover structures in data without accompanying it with an interpretation. Generally, cluster analysis discovers structures in data, but the technique does not explain why those structures exist. Consider a case where a cluster analysis is performed in a market research, researchers can identify specific groups within a population. The analysis of the selected groups can be used to determine the likelihood of a population cluster to purchase particular products or services (Punj & Stewart, 1983). Defining these clusters can help marketing teams to establish targeted advertisements and communication to potential customers.
In summary cluster analysis is beneficial to researchers as they can use the technique to identify and define various patterns within data elements. Reveling patterns existing between data points ensures researchers can outline structures which might have not been identified in previous stages of a research. The structures can portray significant meanings to data once they are discovered.
References
Punj, G. & Stewart, W.D. (1983). Cluster analysis in marketing research: Review and
suggestions for application. Journal of Marketing Research 20(2).
Setyaningsih, S. (2012). Using cluster analysis study to examine the successful performance
entrepreneur in Indonesia. Procedia Economics and Finance, 4, 286-298.
https://doi.org/10.1016/S2212-5671(12)00343-7