Statistical Inferences on Type I and Type II Triangular Distributions
Prof. Dr. Ismail ERDEM

Abstract
If continuous random variables X has a triangular distribution and if its boundary values are unknown, then it may well be necessary to make some statistical inferences related to these parameters. For triangular distributions, with unknown boundary values, some unbiased estimators are proposed. The proposed estimators are compared based on their efficiencies and the best estimators are determined and used for further inferences, such as constructing a confidence interval for , and the tests of hypotheses. By means of a simulation process, matching accuracy between sampling results and theoretical findings is observed.

Full text: PDF