Measuring the complexity of classification problems
- 11 November 2002
- proceedings article
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 2, 43-47
- https://doi.org/10.1109/icpr.2000.906015
Abstract
We studied a number of measures that characterize the difficulty of a classification problem. We compared a set of real world problems to random combinations of points in this measurement space and found that real problems con- tain structures that are significantly different from the ran- dom sets. Distribution of problems in this space reveals that there exist at least two independent factors affecting a prob- lem's difficulty, and that they have notable joint effects. We suggest using this space to describe a classifier's domain of competence. This can guide static and dynamic selection of classifiers for specific problems as well as subproblems formed by confinement, projections, and transformations of the feature vectors.Keywords
All Related Versions
This publication has 10 references indexed in Scilit:
- Meta analysis of classification algorithms for pattern recognitionPublished by Institute of Electrical and Electronics Engineers (IEEE) ,1999
- Large-scale simulation studies in image pattern recognitionPublished by Institute of Electrical and Electronics Engineers (IEEE) ,1997
- An overtraining-resistant stochastic modeling method for pattern recognitionThe Annals of Statistics, 1996
- On the nonlinearity of pattern classifiersPublished by Institute of Electrical and Electronics Engineers (IEEE) ,1996
- An evaluation of intrinsic dimensionality estimatorsPublished by Institute of Electrical and Electronics Engineers (IEEE) ,1995
- An Introduction to Kolmogorov Complexity and Its ApplicationsPublished by Springer Nature ,1993
- Small sample size effects in statistical pattern recognition: recommendations for practitionersPublished by Institute of Electrical and Electronics Engineers (IEEE) ,1991
- A test to determine the multivariate normality of a data setIEEE Transactions on Pattern Analysis and Machine Intelligence, 1988
- Multivariate Generalizations of the Wald-Wolfowitz and Smirnov Two-Sample TestsThe Annals of Statistics, 1979
- Pattern Classifier Design by Linear ProgrammingIEEE Transactions on Computers, 1968