Title :
Qualitativization of 3D Functions from a Discrete Numerical Representation
Author :
Flores, Juan J. ; Rodriguez, Horacio
Author_Institution :
Fac. de Ing. Electr., Univ. Michoacana, Morelia, Mexico
Abstract :
In this paper we propose a qualitative representation for 3D algorithms that given a numeric discrete sampling of a function, produce a qualitative version of it, expressed in the proposed representation. The data may present noisy due to errors, numerical imprecision, etc. We start filtering the information, then we segment the information into monotonic regions. Finally a landmark refinement is needed for generating the qualitative representation of the functions. The generated qualitative representation may then be used to perform qualitative reasoning about the function. The system has been tested by providing information to a program that performs qualitative reasoning about bifurcation diagrams of dynamic systems.
Keywords :
bifurcation; functions; inference mechanisms; information filtering; knowledge representation; 3D function qualitativization; bifurcation diagrams; discrete numerical representation; dynamic systems; information filtering; landmark refinement; numeric discrete sampling; qualitative reasoning; qualitative representation; Artificial intelligence; Bifurcation; Helium; Information filtering; Information filters; Performance evaluation; Physics; Sampling methods; System testing; dynamic systems; qualitative reasoning; qualitative representation;
Conference_Titel :
Artificial Intelligence, 2009. MICAI 2009. Eighth Mexican International Conference on
Conference_Location :
Guanajuato
Print_ISBN :
978-0-7695-3933-1
DOI :
10.1109/MICAI.2009.32
