Rud O.P. Data Mining Cookbook, Modeling Data for Marketing, Risk, and Customer Relationship Management

John Wiley & Sons, Inc., 2001. - 429 p.
Part one: planning the menu
Chapter: setting the objective
Defining the goal
Profile analysis
Segmentation
Response
Risk
Activation
Cross-sell and up-sell
Attrition
Net present value
Lifetime value
Choosing the modeling methodology
Linear regression
Logistic regression
Neural networks
Genetic algorithms
Classification trees
The adaptive company
Hiring and teamwork
Product focus versus customer focusChapter: selecting the data sources
Types of data
Sources of data
Internal sources
External sources
Selecting data for modeling
Data for prospecting
Data for customer models
Data for risk models
Constructing the modeling data set
How big should my sample be?
Sampling methods
Developing models from modeled data
Combining data from multiple offersPart two: the cooking demonstration
Chapter: preparing the data for modeling
Accessing the data
Classifying data
Reading raw data
Creating the modeling data set
Sampling
Cleaning the data
Continuous variables
Categorical variablesChapter: selecting and transforming the variables
Defining the objective function
Probability of activation
Risk index
Product profitability
Marketing expense
Deriving variables
Summarization
Ratios
Dates
Variable reduction
Continuous variables
Categorical variables
Developing linear predictors
Continuous variables
Categorical variables
Interactions detectionChapter: processing and evaluating the model
Processing the model
Splitting the data
Method: one model
Method: two models — response
Method: two models — activation
Comparing method and methodChapter: validating the model
Gains tables and charts
Method: one model
Method: two models
Scoring alternate data sets
Resampling
Jackknifing
Bootstrapping
Decile analysis on key variablesChapter: implementing and maintaining the model
Scoring a new file
Scoring in-house
Outside scoring and auditing
Implementing the model
Calculating the financials
Determining the file cut -off
Champion versus challenger
The two -model matrix
Model tracking
Back-end validation
Model maintenance
Model life
Model logPart three: recipes for every occasion
Chapter: understanding your customer: profiling and segmentation
What is the importance of understanding your customer?
Types of profiling and segmentation
Profiling and penetration analysis of a catalog company's
CustomersRfm analysis
Penetration analysis
Developing a customer value matrix for a credit
Card company
Customer value analysis
Performing cluster analysis to discover customer segmentsChapter: targeting new prospects: modeling response
Defining the objective
All responders are not created equal
Preparing the variables
Continuous variables
Categorical variables
Processing the model
Validation using boostrapping
Implementing the modelChapter: avoiding high-risk customers: modeling risk
Credit scoring and risk modeling
Defining the objective
Preparing the variables
Processing the model
Validating the model
Bootstrapping
Implementing the model
Scaling the risk score
A different kind of risk: fraudChapter: retaining profitable customers: modeling churn
Customer loyalty
Defining the objective
Preparing the variables
Continuous variables
Categorical variables
Processing the model
Validating the model
Bootstrapping