Natthinee Deetae. The classification using empirical bayes in combination with nearest neighbor. Doctoral Degree(Applied Statistics). King Mongkut's University of Technology North Bangkok. Central Library. : King Mongkut's University of Technology North Bangkok, 2013.
The classification using empirical bayes in combination with nearest neighbor
Abstract:
This thesis is aimed to propose a new classification technique using Empirical Bayes in combination with Nearest Neighbor (EBNN) when data are distributed as normal, short-tailed symmetric, long-tailed symmetric, short-tailed asymmetric, long-tailed asymmetric and heavy-tailed distributions. The study is performed using conjugate priors, normal distributions with unknown mean but known variance, and known mean but unknown variance. Data are divided equally into training set and test set with the sample sizes 100, 200 and 500 for the binary classification.In each situation, the data are simulated with Monte Carlo technique and repeated 1,000 times using MATLAB program. The average percentage of correct classification is used as a criterion for comparison.The results of this study indicate that,for 3 levels of sample sizes,Empirical Bayes and EBNN methods tend to yield good classification in all types of data. Moreover, the EBNN method exhibited an improved performance over Empirical Bayes method in all situations under study.