User's Manual

39
Understanding Data Mining
Segmentation nodes
The Auto Cluster node estimates and compares clustering models, which identify
groups of records that have similar characteristics. The node wor ks in the same
manner as other automated m odeling nodes, allowin g you to experiment with multiple
combinations of options in a single modeling pass. Models can be compared using
basic measures w i t h which to attempt to filter and rank the usefulness of the cluster
models, and provide a measure based on the importance of particular fields.
The K-Means node clusters the data set into distin ct groups (or clusters) . The method
defines a fixed number of clusters, iteratively assigns records to clusters, and adjusts
the cluster centers until further r efinement can no longer improve the model. Instead
of trying to predict an outcome, k-means uses a process known as unsupervised
learning to uncover patterns in the set of i nput fields .
The Kohonen node generates a type of neural network that can be used to cluster the
data set into distinct groups. When the network is f ully trained, records that are
similar should be close together on the output map, while records that are different
will be far a part. You can look at the number of observat i ons captured by each unit
in the model nugget to identify the strong units. This may gi ve you a sense of the
appropriate number of clusters.
The TwoStep n ode uses a two-step clustering method. The first step makes a single
pass through the data to compress the raw input data into a manageable set of
subclusters. The second step uses a hierarchical clustering method to progressively
merge the subclusters into larger and larger clusters. TwoStep has the advantage of
automatical l y estimating the optimal number of clusters for the training data. It can
handle mixed field types and large data sets efficiently.
The Anomaly Detection node identifies unusual cases, or outliers, that do not conform
to patterns of “normal” data. With this node, it is possible to identify outliers even if
they do not fit any previou sly known patterns and even if you are not exactly sure
what you are lo ok ing for.
In-Database Mining Models
SPSS Modeler sup ports integration with data mining and modelin g tools that ar e available from
database ve ndors, including Ora cle Data Miner, IBM DB2 Inf oSphere Warehouse, and Microsoft
Analysis Services. You can build, score, and store models inside the database—all from within the
SPSS Modeler application. For full details, see the SPSS Modeler In-Database Mining Guide,
availabl e on the product DVD.
IBM SPSS Statistics Models
If you have a co py of IBM® SPSS® Statistics installed and licens ed on your computer, you can
access and run certain SPSS Statistics routines from within SPSS Modeler to build and score
models .
Further Information
Detailed documentation on the modeling algorithms is also available. For mor e informatio n, see
the SPSS Modeler Algorithms Guide, availab le on the produ ct DVD.