Advanced Search

Browse by Discipline

Scientific Societies

E-print Alerts

Add E-prints

E-print Network

  Advanced Search  

Lehrstuhl fr Informatik 9 Zeit: Dienstag, 13. September 2011, 16.00 Uhr

Summary: Lehrstuhl für Informatik 9
Zeit: Dienstag, 13. September 2011, 16.00 Uhr
Ort: Room 5056, Ahornstr. 55
Referent: Prof. Michael E. Houle
National Institute for Informatics, Tokyo, Japan
Titel: Intrinsic Dimensionality and its Applications to
Databases and Data Mining
In recent years, important applications in the areas of machine learning, data mining,
and bioinformatics have been developed that rely on the search for similarities
among data structures as a fundamental operation. Typically, the computation of
structural similarity is performed using a computationally-expensive exhaustive
search over all possibilities. Effective indexing methods for structural similarity
search, together with appropriate measures of inter-structure similarity, could poten-
tially have a significant impact on the overall performance of such applications.
In this presentation, we propose a general framework within which the effectiveness
of similarity search and other data operations can be assessed, independently of the
nature of the data being considered. Within this framework, we discuss how a form of
intrinsic dimensionality in the neighborhood of a test or query item, the "stereological


Source: Ábrahám, Erika - Fachgruppe Informatik, Rheinisch Westfälische Technische Hochschule Aachen (RWTH)


Collections: Computer Technologies and Information Sciences