Graphical Evaluation of Probabilities of Misclassifications for Normal and Edgeworth Series Distributions
- September 10, 2020
- Posted by: RSIS Team
- Categories: IJRIAS, Mathematics
International Journal of Research and Innovation in Social Science (IJRISS) | Volume V, Issue IV, April 2020 | ISSN 2454–6186
Awogbemi, Clement Adeyeye
Department of Statistics, National Mathematical Centre, Abuja, Nigeria
Abstract:- Failure to transform to normality before classification affects probabilities of misclassification while comparing the distribution of errors of misclassification. A uniformly distributed random variable generated by a varied and repeated method was employed to generate the errors of misclassification for normal and Edgeworth Series Distributions. The proposed method was proved and on the basis of this, an algorithm was developed. There is a non linear dependence of the total probabilities of misclassification on the skewness factor. It was observed that there is a disordered relationship between the probabilities of misclassication for normal and Edgeworth Series Distributions.
Keywords: Graphical Evaluation, Probabilities of Misclassification, Normal Distribution, Edgeworth Series Distribution, Skewness Factor