This preview shows page 1. Sign up to view the full content.
Unformatted text preview: k's customer database into several groups based on similarities
of their age, sex, income, and residence. Once clustering is done, the bank
management can examine the summarized characteristics of each subgroup,
extract significant statistics, and use them to understand its customers better and
thus provide more suitable products and customized services. Clustering
-algorithms have also been successfully applied in the area of text mining for
problems such as document indexing and topic identification in the research
The objective of association algorithms for data mining is to automatically
discover association rules hidden in a database. An association rule reveals the
associative relationships among objects. For example, an association rule may
reveal that "If some object X is part of a transaction, then for some percent of such
transactions, object Y is also part of the transaction." Thus association algorithms
help in the discovery of togetherness, or the connection of objects. They are useful
in solving problems where it is important to understand the extent to which the
presence of some variables imply the presence of other variables and the
prevalence of this pattern in the entire database under examination.
An interesting and very useful application of association algorithm is marketbasket analysis, in which database of sales transactions are examined to extract
patterns that identify what items sell together, what items sell better when
relocated to new areas, and what product groupings improve department sales. For
example, a retail store may discover that people tend to buy soft drinks and potato
chips together. Store personnel then place the potato chips near the soft drinks to
promote the sale of both. They may even discount one to entice buying the other,
since these buyers now will be saving some money.
Another useful application of association algorithm is the analysis of Internet
traffic on a web site to better understand where the real demand is, what pages are
being looked collectively, and so on...
View Full Document
This document was uploaded on 04/07/2014.
- Spring '14