book

Customer Segmentation and Clustering Using SAS Enterprise Miner, Second Edition, 2nd Edition

by Randy Collica

November 2011

Beginner to intermediate

378 pages

10h 48m

English

SAS Institute

Read now

Unlock full access

1.1 - What is Segmentation in the Context of CRM?1.2 - Types of Segmentation and Methods1.2.1 - Customer Profiling1.2.2 - Customer Likeness Clustering1.2.3 - RFM Cell Classification Grouping1.2.4 - Purchase Affinity Clustering1.3 - Typical Uses of Segmentation in Industry1.4 - Segmentation as a CRM Tool1.5 - References
2.1 - Mass Customization Instead of Mass Marketing2.2 - Specialized Promotions or Communications by Segment Groups2.3 - Profiling of Customers and Prospects2.3.1 - Example 2.1: The Data Assay Project2.3.2 - Example 2.2: Customer Profiling of the Buytest Data Set2.3.3 - Additional Exercise2.4 - References
3.1 - What is Similar and What is Not3.2 - Distance Metrics as a Measure of Similarity and Association3.3 - What is Clustering? The k-Means Algorithm and Variations.3.3.1 - Variations of the k-Means Algorithm3.3.2 - The Agglomerative Algorithm3.4 - References
4.1 - Introduction to Cell-Based Segmentation4.2 - Segmentation Using Cell Groups—RFM4.2.1 - Other Cell Types for Segmentation4.3 - Example Development of RFM Cells4.4 - Tree-Based Segmentation Using RFM4.5 - Using RFM and CRM—Customer Distinction4.6 - Additional Exercise4.7 - References4.8 - Additional Reading
5.1 - Motivation for Clustering of Customer Attributes: Beginning CRM5.2 - How Can i Better Understand My Customer Base of Over 100,000?5.3 - Using a Decision Tree to Create Cluster Segments5.4 - References5.5 - Additional Reading
6.1 - Closer to Reality of Customer Segmentation6.2 - Representing Many Attributes in Multi-dimensions6.3 - How Can i Better Understand My Customers of Many Attributes?6.4 - Data Assay and Profiling6.5 - Understanding What the Cluster Segmentation Found6.6 - Planning for Customer Attentiveness with Each Segment6.7 - Creating Cluster Segments on Very Large Data Sets6.8 - Additional Exercise6.9 - References
7.1 - What is the Shelf Life of a Model, and How Can it Affect Your Results?7.2 - How to Detect When your Clustering Model Should be Updated7.3 - Testing New Observations and Score Results7.4 - Other Practical Considerations7.5 - Additional Reading
8.1 - The Basis of Breaking up the Data Space8.2 - Predicting a Segment Level8.3 - Using the Segment Level Predictions for Customer Scoring8.4 - Creating Customer Value Segments8.5 - References8.6 - Additional Exercises
9.1 - Missing Data and How it Can Affect Clustering9.2 - Analysis of Missing Data Patterns9.3 - Effects of Missing Data on Clustering.9.4 - Methods of Missing Data Imputation9.5 - Obtaining Confidence Interval Estimates on Imputed Values9.6 - Using the SAS Enterprise Miner Imputation Node9.7 - References
10.1 - Motivation of Estimating Product Affinity by Segment10.2 - Estimating Product Affinity Using Purchase Quantities10.3 - Combining Product Affinities by Cluster Segments10.4 - Pros and Cons of Segment Affinity Scores.10.5 - issues with Clustering Non-normal Quantities10.6 - Approximating a Graph-Theoretic Approach Using a Decision Tree10.7 - Using the Product Affinities for Cross-Sell Programs10.8 - Additional Exercises10.9 - References
11.1 - When Ordinary Clustering Does not Produce Desired Results11.2 - What is a Self-Organizing Map?11.3 - Computing and Applying SOM Network Cluster Segments11.4 - Comparing Clustering with SOM Segmentation11.5 - Customer Distinction Analysis Example11.6 - Additional Exercises11.7 - References
12.1 - Background of Textual Data in the Context of CRM12.2 - Notes on Text Mining versus Natural Language Processing12.3 - Simple Text Mining Example12.4 - Text Document Clustering12.5 - Using Text Mining in CRM Applications12.6 - References
13.1 - What is Association Analysis and its Uses in Business?13.2 - Market Basket Association Analysis13.3 - Revisiting Product Affinity Using Clustered Associations13.4 - The Business and Technical Side of Clustering Associations.13.5 - References
14.1 - Typical Market Research Surveys.14.2 - Match-back of Survey Responses14.3 - Analysis of Survey Responses: An Overview14.4 - Developing a Predictive Segmentation Model from a Survey Analysis14.5 - Issues with Scoring a Predictive Segmentation on Customer or Prospect Data14.6 - Assessing the Confidence of Predicted Segments14.7 - Business Implications for Using Attitudinal Segmentation.14.8 - References
15.1 - Survey of Methods of Ensemble Segmentations15.2 - Two Methods for Combining Attitudinal and Behavioral Segments15.3 - Presenting the Business Case Simply from a Complex Analysis15.4 - References15.5 - Additional Exercise
16.1 - Measuring Transactions as a Time Series.16.2 - References16.3 - Additional Reading16.4 - Additional Exercise

Content preview from Customer Segmentation and Clustering Using SAS Enterprise Miner, Second Edition, 2nd Edition

Combining Attitudinal and Behavioral Segments

15.1 Survey of Methods of Ensemble Segmentations.

15.2 Two Methods for Combining Attitudinal and Behavioral Segments.

15.3 Presenting the Business Case Simply from a Complex Analysis.

15.4 References.

15.5 Additional Exercise.

15.1 Survey of Methods of Ensemble Segmentations

The word ensemble means to combine, collect, or collaborate. Ensemble models have been around for quite some time. Typical methods for combining different models of the same target variable have been reported as bagging and boosting. Bagging stands for bootstrap aggregation, and one of the first reported bagging algorithms ...