Estimation of the Ld100p When p Is Small

This thesis concerns the estimation of extreme quantiles on a dose-response curve. It focuses on the Robbins-Monro and up-and-down procedures. Simulation studies run in search of the LD15 and LD30 using a variety of methods suggest that the Robbins-Monro procedure is optimal in terms of Monte Carlo MSE and bias. The up-and-down procedure's performance differs in many cases only slightly from that of the Robbins-Monro process, therefore indicating its value as a practical alternative to the Robbins-Monro process for extreme-quantile estimation.

Identiferoai:union.ndltd.org:vcu.edu/oai:scholarscompass.vcu.edu:etd-5434
Date01 January 1997
CreatorsBettendorf, Daniel
PublisherVCU Scholars Compass
Source SetsVirginia Commonwealth University
Detected LanguageEnglish
Typetext
Formatapplication/pdf
SourceTheses and Dissertations
Rights© The Author

Page generated in 0.0019 seconds