Trinity College Dublin, Department of Computer Science
Tsymbal, Alexey. 'Dynamic Integration of Classifiers in the Space of Principal Components'. - Dublin, Trinity College Dublin, Department of Computer Science, TCD-CS-2003-30, 2003, pp15
Computer Science Technical Report TCD-CS-2003-30
Recent research has shown the integration of multiple classifiers to
be one of the most important directions in machine learning and data mining. It
was shown that, for an ensemble to be successful, it should consist of accurate
and diverse base classifiers. However, it is also important that the integration
procedure in the ensemble should properly utilize the ensemble diversity. In this
paper, we present an algorithm for the dynamic integration of classifiers in the
space of extracted features (FEDIC). It is based on the technique of dynamic
integration, in which local accuracy estimates are calculated for each base
classifier of an ensemble, in the neighborhood of a new instance to be
processed. Generally, the whole space of original features is used to find the
neighborhood of a new instance for local accuracy estimates in dynamic
integration. In this paper, we propose to use feature extraction in order to cope
with the curse of dimensionality in the dynamic integration of classifiers. We
consider classical principal component analysis and two eigenvector-based
supervised feature extraction methods that take into account class information.
Experimental results show that, on some data sets, the use of FEDIC leads to
significantly higher ensemble accuracies than the use of plain dynamic
integration in the space of original features. As a rule, FEDIC outperforms plain
dynamic integration on data sets, on which both dynamic integration works (it
outperforms static integration), and considered feature extraction techniques are
able to successfully extract relevant features.
Please note: There is a known bug in some browsers that causes an
error when a user tries to view large pdf file within the browser window.
If you receive the message "The file is damaged and could not be
repaired", please try one of the solutions linked below based on the
browser you are using.
Items in TARA are protected by copyright, with all rights reserved, unless otherwise indicated.