|
Foreword by Donald K. Burleson
Oracle Data Mining and Predictive Analytics
Why this book is important
Chapter 1: Introduction to Model Building
What is Data Mining?
Components of Oracle Data Miner
Sampling Data from the Database
Concentrating on a customer
Building a Classification Model
Naming Data Mining Activities
Running a Data Mining Activity
Viewing your Results
The ODM ROC Curve
Applying changes to a Model
Attribute Importance in the Naïve Bayes Model
Building Naïve Bayes Model with Fewer Attributes
Applying the Model
Using the Create View Wizard
Scoring New Data
Viewing Top Rankings
Conclusion
Chapter 2: Adaptive Bayes Network and Decision Trees
Introduction to Classification
Data Mining Classification Models
Using the Models
Importing a Dataset
Exploring and Reducing the Dataset
Viewing Attribute Histograms
Attribute Importance
Comparing Naïve Bayes Models for Forest Cover
Adaptive Bayes Single Feature Model
Building the Adaptive Bayes Network Model
Sampling
Viewing Adaptive Bayes Network Results
Interpreting Adaptive Bayes Network Results
Building the Adaptive Bayes Multi Feature Model
Using the ROC Feature
Introducing Cost Bias to the Classification Model
Building a Decision Tree
The Decision Tree Classification Model
Decision Tree Classification Rules
Conclusion
Chapter 3: Using Support Vector Machines
Introduction to Support Vector Machine
Inside Support Vector Machines
Importing the Irish Wind Data File
Computing a New Attribute with Compute Field
Transformation Wizard
Building the SVM Model
Handling Outlier Values in SVM Analysis
Missing Values in SVM Analysis
Sparse Data in SVM Analysis
Normalization of SVM Data
Linear and Gaussian Kernels
SVM and Over-fitting
SVM Results with Gaussian Kernel
Importing Boston House Price Data
Building SVM Classification Models
Interpreting the SVM Results
Refining the SVM Model
Building a SVM Regression Model
Regression Model Results
Linear Regression Analysis
Drilling into the SVM Data
Using Text Data in SVM Predictive Models
Importing CLOB Data
Loading CLOB Data into the Oracle Database
Building a SVM Text Model
Interpreting the SVM text Data
Conclusion |
Chapter 4: Creating Clusters and Cohorts
Clustering and Cohorts
The k-Means Cluster
Using O-Cluster
O-Cluster Sensitivity Settings
Using K-Means for Clustering
Examining the CoIL Data
Building a K-Means Cluster
Finding majority cohort values
Comparing data sub-sets with K-Means
Choosing the Appropriate Data Mining Algorithm
When to use K-Means Analysis
When to use O-Cluster Analysis
Applying the Cluster
Publishing the Cluster Results
Publishing to a File
Using the Discoverer Gateway for Publication
Publishing to an Oracle Database
Importing the model to a different Oracle database
Conclusion
Chapter 5: Inside Oracle Data Miner
Exploring Data Miner
Data Miner Activity Builder Tasks
Quantile Binning
Using the Discretize Transform Wizard
Customizing Discretize Transformations
Using the Aggregate Transformation Wizard
Recode Transformation Wizard
Using the Split Transformation Wizard
Using the Stratified Sample Transformation Wizard
Using the Filter Single-Record Transformation Wizard
Inside the Sample Transformation Wizard
Preparing datasets for Data Mining Activities
Using the Missing Values Transformation Wizard
Using the Normalize Transformation Wizard
Using the Numeric Transformation Wizard
Using the Outlier Treatment Transformation Wizard
Conclusion
Chapter 6: Predictive Analytics
Predictive Analytics in Data Mining
Explain Procedure
Predict Procedure
Explain Wizard
Predict Wizard
Applying Predictive Analytics
Conclusion
Chapter 7: Personalized Form Letter Generation with
Oracle BI Publisher
Scenarios for using ODM with BI Publisher
Building a Decision Tree Model
Results of the Decision Tree Model
Scoring the Apply Dataset.
Using SQL to View Results of Scored Data
Creating a Report using BI Publisher Enterprise Server
Using Template Builder for Oracle BI Publisher
Adding Fields to the Word Template using BI Publisher
Template Builder
Creating a Personalized Customer Letter with Three
Offers
Scenario for Personalizing a Form Letter
Building a Decision Tree Model using Oracle Data Miner
Accuracy of the Fund Raiser DT Model
Results of the Fund Raiser DT Model
Generating XML Data using BI Publisher
Creating a Form Letter with the Template Builder
Conclusion
Book Conclusion
Appendix A: Installing Oracle Data Miner
ODM Tutorial
Purpose
Time to Complete
Topics
Overview
Prerequisites
Enabling the DMSYS Account
Creating and Configuring A Data Mining Account
Installing Oracle Data Miner
Summary
Appendix B: Script to Create ODM User
Scripts |
