Data Preparation for Analytics Using SAS
- 5h 48m
- Gerhard Svolba
- SAS Institute
- 2006
Written for anyone involved in the data preparation process for analytics, this user-friendly text offers practical advice in the form of SAS coding tips and tricks, along with providing the reader with a conceptual background on data structures and considerations from the business point of view. Topics addressed include viewing analytic data preparation in the light of its business environment, identifying the specifics of predictive modeling for data mart creation, understanding the concepts and considerations for data preparation for time series analysis, using various SAS procedures and SAS Enterprise Miner for scoring, creating meaningful derived variables for all data mart types, using powerful SAS macros to make changes among the various data mart structures, and more!
In this Book
-
Analytic Business Questions
-
Characteristics of Analytic Business Questions
-
Characteristics of Data Sources
-
Different Points of View on Analytic Data Preparation
-
The Origin of Data
-
Data Models
-
Analysis Subjects and Multiple Observations
-
The One-Row-per-Subject Data Mart
-
The Multiple-Rows-per-Subject Data Mart
-
Data Structures for Longitudinal Analysis
-
Considerations for Data Marts
-
Considerations for Predictive Modeling
-
Accessing Data
-
Transposing One- and Multiple-Rows-per-Subject Data Structures
-
Transposing Longitudinal Data
-
Transformations of Interval-Scaled Variables
-
Transformations of Categorical Variables
-
Multiple Interval-Scaled Observations per Subject
-
Multiple Categorical Observations per Subject
-
Coding for Predictive Modeling
-
Data Preparation for Multiple-Rows-per-Subject and Longitudinal Data Marts
-
Sampling
-
Scoring and Automation
-
Do's and Don'ts When Building Data Marts
-
Case Study 1—Building a Customer Data Mart
-
Case Study 2—Deriving Customer Segmentation Measures from Transactional Data
-
Case Study 3—Preparing Data for Time Series Analysis
-
Case Study 4—Data Preparation in SAS Enterprise Miner
