Advanced Analytics for Customer Intelligence Using SAS

Duration: 3.0 days

This advanced, highly interactive course will clarify how you can adopt state-of-the-art data mining techniques for complex customer intelligence applications. You will receive a sound mix of both theoretical and technical insights as well as practical implementation details, illustrated by several real-life cases.

Learn how to

• apply a series of powerful, recently developed, cutting-edge data mining techniques
• ensure the practical application of these techniques to optimize strategic business decisions
• explore a futuristic vision of how new emerging data mining techniques may change your key business processes
• deploy, monitor, and optimally back-test data mining systems.

Who should attend: Those involved in estimating, monitoring, or maintaining predictive models for various types of customer intelligence

Prerequisites

Before attending this course, you should know how to

• preprocess data such as missing values, outliers, categorization, and sampling
• develop predictive models using logistic regression
• develop predictive models using decision trees
• develop descriptive models using basic segmentation techniques
• quantify the performance of predictive models such as lift curves and ROC curves.

Course Contents

Introduction

• customer intelligence: basic nomenclature and a review
• predictive modeling for customer intelligence: the KDD process model
• refresher on data preprocessing
• refresher on basic predictive modeling techniques (for example: logistic regression, decision trees, and k-nearest neighbor)

Advanced Sampling Schemes

• cross-validation (stratified, leave-one-out)
• bootstrapping

Neural Networks

• multilayer perceptrons (MLPs)
• MLP types (RBF, recurrent, etc.)
• overfitting and weight regularization
• Hinton graphs for input selection
• self-organizing maps (SOMs)
• case study: SOMs for market segmentation

Support Vector Machines (SVMs)

• the kernel trick and Mercer's theorem
• SVMs for classification and regression
• hyperparameter tuning using cross-validation methods
• case study: SVMs for response modeling

Rule Extraction from Neural Networks and SVMs

• turning black-box models into white-box models
• pedagogical versus decompositional approaches

Regression Trees

• splitting/stopping/assignment criteria

Ensemble Methods

• bagging and boosting
• stacking
• random forests
• case study: ensemble methods for fraud detection

Alternative Rule Representation Formats

• decision tables
• decision graphs

Bayesian Network Classifiers

• naive Bayes
• tree augmented naive Bayes
• unrestricted Bayesian network classifiers
• Bayesian inference
• case study: Bayesian network inference for churn prediction

Survival Analysis

• Kaplan Meier analysis
• parametric survival analysis
• proportional hazards regression
• neural networks for survival analysis
• case study 1: customer lifetime value modeling using survival analysis
• case study 2: credit scoring using survival analysis

Other Predictive Modeling Techniques

• genetic algorithms
• fuzzy techniques
• ant colony optimization

Multi-Relational Data Mining

• relational versus flat data
• inductive logic programming
• relational decision trees, relational association rules, etc.

Mining Networked Data

• social network mining
• network autocorrelation
• univariate network classification techniques (guilt-by-association)
• social network marketing
• case study: CRM applications of network models

Monitoring and Backtesting Advanced CRM Models

• traffic light indicator approaches
• population stability
• model stability
• impact of macro-economic effects
• case study: monitoring and backtesting a credit scoring model

Software Addressed

This course addresses the following software product(s): SAS Enterprise Miner.

This page was created using SAS software.