Data Mining Overview Lecture Objectives After this lecture

Data Mining Overview

Lecture Objectives After this lecture, you should be able to: 1. Explain key data mining tasks in your own words. 2. Discuss one broad business application of data mining. 3. Explain one way to evaluate effectiveness of a Data Mining project.

Data Mining Tasks 1. Description/Visualization n 2. Segmentation n 3. Regression Techniques – Linear, Logistic Association n 5. Cluster Analysis Prediction / Classification n 4. Charts/Graphs/Tabulations Market Basket Analysis Optimization n Linear Programming

Course Overview/Techniques Used Data Preparation Prediction/Classification Linear Classification (Discriminant Analysis) Classification Trees (CART, CHAID) Logistic Regression Artificial Neural Networks Segmentation Judgment Cluster Analysis Factor Analysis Association Matching techniques Market Basket Analysis

Application in Financial Services Stage 1 Product Planning Customer Stage 2 Acquisition Customer Valuation Stage 4 Collections and Recovery Customer Management Stage 3

Measuring Effectiveness: Percent of potential responders captured Lift/Gains Chart 100 Targeting 90 Random mailing 45 0 45 100 Percent of population targeted Dr. Satish Nargundkar

Discussion 1. Can you think of other applications? 2. What are some limitations of Data Mining? 3. What are future possibilities?