Introduction to SAS Enterprise Miner 5.3 Software

Example Problem Description

A national charitable organization seeks to better target its solicitations for donations. By only soliciting the most likely donors, less money will be spent on solicitation efforts and more money will be available for charitable concerns. Solicitations involve sending a small gift to an individual along with a request for a donation. Gifts include mailing labels and greeting cards.

The organization has more than 3.5 million individuals in its mailing database. These individuals have been classified by their response to previous solicitation efforts. Of particular interest is the class of individuals who are identified as lapsing donors. These individuals have made their most recent donation between 12 and 24 months ago. The organization has found that by predicting the response of this group, they can use the model to rank all 3.5 million individuals in their database. The campaign refers to a greeting card mailing sent in June of 1997. It is identified in the raw data as the 97NK campaign.

When the most appropriate model for maximizing solicitation profit by screening the most likely donors is determined, the scoring code will be used to create a new score data set that is named Donor.ScoreData. Scoring new data that does not contain the target is the end result of most data mining applications.

When you are finished with this example, your process flow diagram will resemble the one shown below.

[untitled graphic]

Here is a preview of topics and tasks in this example:

Chapter	Task
2	Create your project, define the data source, configure the metadata, define prior probabilities and profit matrix, and create an empty process flow diagram.
3	Define the input data, explore your data by generating descriptive statistics and creating exploratory plots. You will also partition the raw data and replace missing data.
4	Create a decision tree and interactive decision tree models.
5	Impute missing values and create variable transformations. You will also develop regression, neural network, and autoneural models. Finally, you will use the variable selection node.
6	Assess and compare the models. Also, you will score new data using the models.
7	Create model results packages, register your models, save and import the process flow diagram in XML.

Note: This example provides an introduction to using Enterprise Miner in order to familiarize you with the interface and the capabilities of the software. The example is not meant to provide a comprehensive analysis of the sample data. [cautionend]

Top of Page