Graduate Student Seminar

Speaker 1: Luan Chip Nguyen

Title

The Metropolis-Hastings Algorithm

Abstract

Monte Carlo methods have played immense role in applied mathematics, statistics and computation physics. One of well-known algorithms for sampling is the Metroplis-Hastings algorithm which is Markov Chain Monte Carlo (MCMC) method to produce random samples from a probability distribution. This talk will introduce the original Metropolis algorithm and its generalization by Hastings. Some of its application will also be discussed.

Speaker 2: Rabab Elnaiem

Title

Statistical Analysis of Bioequivalence

Abstract

For the approval of generic drug products, the commonly used method is to assess average bioequivalence (ABE) in terms of drug absorption through the conduct of bioequivalence studies, assessing if the rate (efficiency) and the extent (safety) of the test drug is “similar” to the reference drug. In bioequivalence studies, the data obtained are the blood or plasma concentrations measured at various time points after the drugs are administered to a sample of subjects using appropriate crossover designs. The pharmacokinetic (PK) responses that are used for assessing bioequivalence are lognormally distributed. In this talk we review the classical two one-sided test for assessing bioequivalence and discuss alternative for when the test is not suitable.