IBM SPSS Modeler 18.3 Algorithms Guide helps organizations to improve customer and citizen relationships through an in-depth understanding of data.
IBM SPSS Modeler is described as an enterprise-strength data mining workbench, and this edition applies to IBM SPSS Modeler 18.3 and subsequent releases. The guide documents modeling algorithms used by the software.
It includes adjusted propensity score calculation methods and the anomaly detection procedure, including notation, algorithm steps, modeling, scoring, and reasoning stages.
It searches for unusual cases based on deviations from the norms of their cluster groups.
The default value is 6.
The neural network uses one hidden layer with 3 neurons.