Class AttributeSelection

  • All Implemented Interfaces:
    adams.core.Destroyable, adams.core.GlobalInfoSupporter, adams.core.logging.LoggingLevelHandler, adams.core.logging.LoggingSupporter, adams.core.option.OptionHandler, adams.core.SizeOfHandler, adams.opt.optimise.FitnessFunction, Serializable

    public class AttributeSelection
    extends AbstractWEKAFitnessFunction
    Perform attribute selection using WEKA classification.
    Version:
    $Revision$
    Author:
    dale
    See Also:
    Serialized Form
    • Constructor Detail

      • AttributeSelection

        public AttributeSelection()
    • Method Detail

      • getDataDef

        public adams.opt.optimise.OptData getDataDef()
      • getWeights

        protected int[] getWeights​(adams.opt.optimise.OptData opd)
      • evaluate

        public double evaluate​(adams.opt.optimise.OptData opd)
      • getRemoveAsString

        public String getRemoveAsString​(int[] m_weights)
        Generates a range string of attributes to keep (= one has to use the inverse matching sense with the Remove filter).
        Returns:
        the range of attributes to keep
      • newBest

        public void newBest​(double val,
                            adams.opt.optimise.OptData opd)
        Callback for best measure so far
        Specified by:
        newBest in interface adams.opt.optimise.FitnessFunction
        Overrides:
        newBest in class adams.opt.optimise.AbstractFitnessFunction