Using KNN to match similar cases
K-Nearest Neighbors (KNN) is found in the Classification tab of the Modeling palette, but it is actually two different applications in one node. Methodologically they are similar, but different in their application. In addition to classification, the KNN Modeling node can be used to calculate differences and find those cases in a Deployment data set that have the smallest distances. Online dating services use this approach. Men can be scored to find women that they resemble, and women can be scored to find men that they resemble. In the case of a dating service, the answers to survey questions define similarity.
In this recipe, we will focus on patterns of purchase. The goal will be to have online sales reps assigned ...
Get IBM SPSS Modeler Cookbook now with the O’Reilly learning platform.
O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.
