| Autor: |
Minkah, Richard, de Wet, Tertius, Ghosh, Abhik, Yousof, Haitham M. |
| Předmět: |
|
| Zdroj: |
Communications for Statistical Applications & Methods; Nov2023, Vol. 30 Issue 6, p531-550, 20p |
| Abstrakt: |
This article presents a new estimator for extreme quantiles in heavy-tailed distributions. The estimator is based on an exponential regression model and is found to have less bias and mean square error compared to existing estimators. The performance of the proposed estimator is evaluated through a simulation study, which shows its robustness in estimating extreme quantiles. The estimator is also applied to practical datasets from the pedochemical and insurance industries, demonstrating its stable and reduced-bias estimates. The article concludes by mentioning potential applications and future research directions for the estimator. [Extracted from the article] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
