Title :
Feature selection using multi-objective genetic algorithms for handwritten digit recognition
Author :
Oliveira, L.S. ; Sabourin, R. ; Bortolozzi, F. ; Suen, C.Y.
Author_Institution :
Ecole de Technologie Superieure, Montreal, Que., Canada
Abstract :
Discusses the use of genetic algorithms for feature selection for handwriting recognition. Its novelty lies in the use of multi-objective genetic algorithms where sensitivity analysis and neural networks are employed to allow the use of a representative database to evaluate fitness and the use of a validation database to identify the subsets of selected features that provide a good generalization. Comprehensive experiments on the NIST database confirm the effectiveness of the proposed strategy.
Keywords :
genetic algorithms; handwritten character recognition; multilayer perceptrons; pattern classification; sensitivity analysis; NIST database; feature selection; handwritten digit recognition; multi-objective genetic algorithms; multi-objective optimization; neural network; representative database; sensitivity analysis; validation database; Algorithm design and analysis; Constraint optimization; Genetic algorithms; Handwriting recognition; Machine intelligence; NIST; Pattern analysis; Pattern recognition; Spatial databases; Vectors;
Conference_Titel :
Pattern Recognition, 2002. Proceedings. 16th International Conference on
Print_ISBN :
0-7695-1695-X
DOI :
10.1109/ICPR.2002.1044794