DataMining-AssociationRulesandClustering.ppt

Data Mining-Association Rules and Clustering Prof. Sin-Min Lee What can be inferred? I purchase diapers I purchase a new car I purchase OTC cough medicine I purchase a prescription medication I don’t show up for class The Introduction to Data Mining The process of extracting valid, previously unknown, comprehensible, and actionable information from large databases and using it to make crucial business decisions. OVERVIEW 1. What is predictive modeling 2. Two phases of predictive modeling 1)Training phase 2)Testing phase 3. Two techniques of predictive modeling 1)Classification Tree induction Neural network 2)Value Prediction Linear regression Nonlinear regression 1. Predictive modeling Similar to human learning experience Use observations ! Form a model of important characteristics of some phenomenon 1. Predictive modeling (contd.) A “black box” that makes predictions about the future based on information from the past and present Application: customer retention management, credit approval, direct marketing. 1. Predictive modeling (contd.) Definitions Maps data item into one of several clusters, where clusters are natural groupings of data items based on similarity metrics or probability density models. Multi-scale representation of data refers to visualization of the data at different ‘scales’, where the term scale may signify either unit, frequency, radius, window size or kernel parameters. Definitions The nontrivial process of identifying valid, novel, potentially useful, and ultimately understandable patterns in data Drawing circles and shapes around dots representing objects! Overview of Clustering Cluster analysis has wide application including market/customer segmentation, pattern recognition, biological studies, and Web document classification Clustering is a dynamic field of research in data mining. The algorithms can be categorized into partitioning, hierarchical, density-based, and model-based methods Data T