KNFST Learner

Calculates a Kernel Null Foley-Sammon Transformation that can be used to calculate a novelty score for a test sample in relation to the training samples.

The corresponding source code is a java adaption of the Matlab code provided by Bodesheim et al. for their paper "Kernel Null Space Methods for Novelty Detection".

Input ports

1. Type: Table
   Training data
   Training data

Output ports

1. Type: Table
   Target points
   Target points. Each class of the input table is represented by one point in the nullspace to which all instances of that class are projected.
2. Type: KNFST Model
   KNFST-model
   KNFST-model that can be used by a KNFST Novelty Scorer node to rate the novelty of instances