Data Mining (Essay Sample)

Essay, Computer sciences and Information technology Data mining
Student Name
Date
Types of Data Mining
Each type or form of data mining has various unique patterns as discussed below.
Class Description
As discussed in the book, Class description identifies the properties of a certain collection of data objects (Abouelhoda & Ghanem 2010, p67). A good example is that of identifying properties of people who travel in smaller vans from those who use large buses. In this example, size is the distinguishing property. Another example is that which separates a group of people wearing red shirts from those wearing green shirts. This example uses color of the shirts as the distinguishing property.
Class Discrimination
Class discrimination as discussed in the book identifies characteristics of different groups of people or objects. It is used to locate the properties, which separate a superior class from the inferior one. A good example is the method that identifies residents of a high-class suburb estate and those of shanty slum areas. Another example is the identification of the students taking their Master’s Degree classes from the undergraduate students.
Cluster Analysis
Cluster analysis involves the grouping of a collection of items to make objects of one group have more resemblance with each other than with the ones on a different group. For example, in studying the social website usage, clustering can apply in the identification of user groups with large number of people. Another example is the division of multimedia data items as sound, graphics and text.
Association Analysis
This is the study of the connections between different groups of data (Omiecinski 2003, p57). For example, when many clients buy items, their information (data) is taken every day at the service points of the retail outlets. This forms the base of all customer transactions. Every record in the database forms a transaction, with a unique transaction identification field. The trader can perform analysis on the data to know the clients trend of buying.
Outlier Analysis
This is the identification of data items, which fall outside the expected range of measurement (Kantardzic 2003, p34). For example, when the ages of members of a group are being analyzed, it is possible to find the age of a person reading 12789 years. This is identified as an error because it is not practically possible to have ages of normal human beings being of that magnitude. Another example is that of various currencies involved in arithmetic computation without conversion. For example, add...