Hitachi Vantara Pentaho Community Wiki
Skip to end of metadata
Go to start of metadata

Package

weka.filters.unsupervised.attribute

Synopsis

A propositionalization filter inspired by the RELAGGS algorithm.
It processes all relational attributes that fall into the user defined range (all others are skipped, i.e., not added to the output). Currently, the filter only processes one level of nesting.
The class attribute is not touched.

For more information see:

M.-A. Krogel, S. Wrobel: Facets of Aggregation Approaches to Propositionalization. In: Work-in-Progress Track at the Thirteenth International Conference on Inductive Logic Programming (ILP), 2003.

Options

The table below describes the options available for RELAGGS.

Option

Description

debug

Turns on output of debugging information.

invertSelection

Set attribute selection mode. If false, only selected attributes in the range will be worked on; if true, only non-selected attributes will be processed.

maxCardinality

The maximum number of values a nominal attribute can have before it's skipped.

Capabilities

The table below describes the capabilites of RELAGGS.

Capability

Supported

Class

Numeric class, Date class, Binary class, Nominal class, No class, Missing class values

Attributes

Empty nominal attributes, Relational attributes, Date attributes, Nominal attributes, Numeric attributes, Binary attributes, Missing values, Unary attributes

Min # of instances

0

  • No labels