SAS Visual Statistics
is an add-on modeling tool to SAS Visual Analytics. The modeling capabilities
of SAS Visual Statistics are provided in SAS Visual Analytics Explorer
(the explorer). The visualizations provided by SAS Visual Statistics
include a linear regression, a logistic regression, a generalized
linear model (GLM), and a clustering tool. In addition, you gain access
to a more powerful decision tree visualization, which includes the
ability to interactively train a decision tree by manually selecting
which branches to prune, split, or train. Model assessment statistics
are available in the core decision tree visualization. To export your
models, you can export model score code for any model or export derived
columns that contain prediction information (predicted values, residuals,
cluster ID, and so on).
Within SAS Visual Statistics,
you can perform a variety of tasks critical to creating good predictive
models. These tasks include data segmentation (supervised and unsupervised),
supervised variable transformation, stratified modeling, outlier detection,
interactive feature creation, data filtering, and post-model visualization.
The examples in this book are intended to familiarize you with the
SAS Visual Statistics interface and workflow. Therefore, not all of
these tasks are covered in this book.