Classification Rules for Exponential Populations Under Order Restrictions on Parameters
Classification procedures of an observation into one of two exponential populations is considered. Assuming a known order between the population parameters, a class of classification rules is proposed. Our study shows that each classification rule in the class is better than the likelihood ratio based classification rule. Comparison of these classification rules with respect to correct probability of classification has been done by extensive simulations.
KeywordsExponential population Classification rule Probability of correct classification
- 2.Barlow, R.E., Bartholomew, D.J., Bremnner, J.M., Brunk, H.D.: Statistical Inference under Order Restrictions. The Theory and Application of Isotonic Regression. John Wiley, New York (1972)Google Scholar
- 3.Basu, A.P., Gupta, A.K.: Classification rules for exponential populations. In: Proschan F., Serfling R.J. (eds.) Reliability and Biometry: Statistical Analysis of Life Length, pp. 637–650. SIAM, Philadelphia (1974)Google Scholar
- 4.Basu, A.P., Gupta, A.K.: Classification Rules for Exponential Populations: Two Parameter Case. DTIC Document, North-Holland (1976)Google Scholar
- 9.Rausand, M., Høyland, A.: System Reliability Theory: Models, Statistical Methods, and Applications. John Wiley, New York (2004)Google Scholar