Small sample size effects in statistical pattern recognition: recommendations for practitioners
- 1 March 1991
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 13 (3) , 252-264
- https://doi.org/10.1109/34.75512
Abstract
Summary:Small learning-set properties of the Euclidean distance, the Parzen window, the minimum empirical error and the nonlinear single layer perceptron classifiers depend on an “intrinsic dimensionality” of the data, however the Fisher linear discriminant function is sensitive to all dimensions. There is no unique definition of the “intrinsic dimensionality”. The dimensionality of the subspace where the data points are situated is not a sufficient definition of the “intrinsic dimensionality”. An exact definition depends both, on a true distribution of the pattern classes, and on the type of the classifier usedKeywords
This publication has 36 references indexed in Scilit:
- Error Rate Estimation in Discriminant Analysis: Recent AdvancesPublished by Springer Nature ,1987
- Assessing the performance of an allocation ruleComputers & Mathematics with Applications, 1986
- 21 Logical functions in the problems of empirical predictionPublished by Elsevier ,1982
- The efficiency of Efron's “Bootstrap” Approach Applied to Error Rate Estimation in Discriminant AnalysisJournal of Statistical Computation and Simulation, 1980
- The General Distribution of the Error Rate of a Classification Procedure with Application to Logistic Regression DiscriminationJournal of the American Statistical Association, 1980
- An Intrinsic Dimensionality Estimator from Near-Neighbor InformationPublished by Institute of Electrical and Electronics Engineers (IEEE) ,1979
- The bias of the apparent error rate in discriminant analysisBiometrika, 1976
- Error estimation in pattern recognition viaL_alpha-distance between posterior density functionsIEEE Transactions on Information Theory, 1976
- Robustness of the linear and quadratic discriminant function to certain types of non‐normalityCommunications in Statistics, 1973
- A note on learning for Gaussian propertiesIEEE Transactions on Information Theory, 1965