ITEC 3040 Study Guide - Final Guide: Euclidean Distance, Kelvin, Covariance
Document Summary
To have predication such as customer loss, we need model first. We can do correlation analysis between each attribute and the class attribute. Clustering analysis is a technique in which we have a dataset, each of the data object is described by set attribute. Data points in set are grouped into different group, within a group the data points will be similar to each other in one cluster but are less similar in separate cluster. The purpose of clustering is that we group data point that are near to each other similar to each other in one group but between the group they are not. In clustering we don"t have class attributes, we don"t know any grouping information. Set of records each of which contains some numbers of items from a given collection, need dependency rule which will predict occurrence of an item based on occurrences of other items. Occurrence of one item can lead to occurrence of other.
