Statistical Methods of Discrimination and Classification: Advances in Theory and Applications is a collection of papers that tackles the multivariate problems of discriminating and classifying subjects into exclusive population. The book presents 13 papers that cover that advancement in the statistical procedure of discriminating and classifying. The studies in the text primarily focus on various methods of discriminating and classifying variables, such as multiple discriminant analysis in the presence of mixed continuous and categorical data; choice of the smoothing parameter and efficiency of k-nearest neighbor classification; and assessing the performance of an allocation rule. The book will be of great use to researchers and practitioners of wide array of scientific disciplines, including engineering, psychology, biology, and physics.
Foreword Discrimination and Classification: Overview Multiple Discriminant Analysis in the Presence of Mixed Continuous and Categorical Data On the Estimation of the Expected Probability of Misclassification in Discriminant Analysis with Mixed Binary and Continuous Variables Parametric and Kernel Density Methods in Discriminant Analysis: Another Comparison Multiple Group Logistic Discrimination Distribution-Free Partial Discrimination Procedures Choice of the Smoothing Parameter and Efficiency of Â:-Nearest Neighbor Classification Monte Carlo Study of Forward Stepwise Discrimination Based on Small Samples The Robust Estimation of Classification Error Rates Assessing the Performance of an Allocation Rule The Variance of the Error Rates of Classification Rules Estimating Class Sizes by Adjusting Fallible Classifier Results On a Classification Rule for Multiple Measurements
- No. of pages:
- © Pergamon 1986
- 12th May 1986
- eBook ISBN: