A Theoretical Analysis of Feature Fusion in Stacked Generalization


Ozay M., YARMAN VURAL F. T.

IEEE 17th Signal Processing and Communications Applications Conference, Antalya, Turkey, 9 - 11 April 2009, pp.774-777 identifier

  • Publication Type: Conference Paper / Full Text
  • City: Antalya
  • Country: Turkey
  • Page Numbers: pp.774-777

Abstract

In the present work, a theoretical framework in order to define the general performance of stacked generalization learning algorithm is developed. Analytical relationships between the performance of the Stacked Generalization classifier relative to the individual classifiers are constructed by the proposed theorems and the practical techniques are developed in order to optimize the performance of stacked generalization algorithm based on these relationships.