In this paper we present a novel method for scoring function specification and feature selection by combining unsupervised learning with supervised cross validation. A one dimensional Kohonen SOM is used to perform a clustering of object-data for a chosen subset of input features and given number of clusters. The resulting object clusters are compared with the predefined original object classes and a matching factor (score) is calculated. This score is used as criterion function for heuristic sequential feature selection.
|Title of host publication||Proceedings of IEEE APCAST'12 Conference|
|Publication status||Published - 2012|
|Event||IEEE APCast'12 - Sydney, Australia|
Duration: 6 Feb 2012 → 8 Feb 2012
|Period||06.02.2012 → 08.02.2012|
- feature selection
- Kohonen SOM