Context Out Classifier

  • Radek Hrebik
  • Jaromir Kukal
Keywords: classification, binary programming, cluster union, imperfect learning


Novel context out learning approach is discussed as possibility of using simple classifiers which is background of hidden class system. There are two ways how to perform final classification. Having a lot of hidden classes we can build their unions using binary optimization task. Resulting system has the best possible sensitivity over all output classes. Another way is to perform second level linear classification as referential approach. The presented techniques are demonstrated on traditional iris flower task.


Boser, B.E., Guyon, I.M., Vapnik, V.N.: A training algorithm for optimal margin classifiers. In: Proceedings of the Fifth Annual Workshop on Computational Learning Theory. pp. 144–152. COLT ’92, New York, NY, USA (1992)

Chang, L., Slikker, W.: Neurotoxicology: Approaches and Methods. Elsevier Science (1995)

Fisher, R.A.: The use of multiple measurements in taxonomic problems. Annals of eugenics 7(2), 179–188 (1936)

Hoerl, A.E., Kennard, R.W.: Ridge regression: Biased estimation for nonorthogonal problems. Technometrics 12(1), 55–67 (1970)

Hrebik, R., Kukal, J.: Application of kohonen som learning in crisis prediction. In: Mathematical Methods in Economics. pp. 254–258. University of Hradec Kralove, Hradec Kralove (2017)

Hrebik, R., Kukal, J., Jablonsky, J.: Optimal unions of hidden classes. Central European Journal of Operations Research pp. 1–17 (2017)

Novakova, K.: Application of Transforms in Object Recognition (in Czech). Ph.D. thesis, FNSPE, CTU in Prague (2008)

Taylor, J.: An Introduction to Error Analysis: The Study of Uncertainties in Physical Measurements. A series of books in physics, University Science Books (1997)

Vittinghoff, E., Glidden, D.V., Shiboski, S.C., McCulloch, C.E.: Regression methods in biostatistics: linear, logistic, survival, and repeated measures models. Springer Science & Business Media (2011)

How to Cite
Hrebik, R. and Kukal, J. 2018. Context Out Classifier. MENDEL. 24, 1 (Jun. 2018), 101-106. DOI: