Online Program Home
My Program

Abstract Details

Activity Number: 540 - SPEED: Clinical Trial Design, Longitudinal Analysis, and Other Topics in Biopharmaceutical Statistics
Type: Contributed
Date/Time: Wednesday, August 1, 2018 : 11:35 AM to 12:20 PM
Sponsor: Biopharmaceutical Section
Abstract #332874
Title: Exposure-Response Analysis with Random Forest
Author(s): Zifang Guo* and Thomas Jemielita and John Kang
Companies: Merck and Merck & Co. and Merck
Keywords: exposure-response analysis; drug development; logistic regression; machine learning; random forest

Over the past decade, exposure-response analysis has become an integral part of clinical drug development and regulatory decision-making. The current practice of exposure-response analysis typically relies on parametric modeling and involves step-wise procedures consisting of structural model selection, covariate selection, model fitting and model prediction. However, this current practice is subject to multiple issues such as model mis-specification and error propagation. In this presentation, we will discuss the application of random forest in exposure-response analysis along with its challenges and solutions. A new method utilizing both random forest and parametric modeling is proposed. Simulation results comparing the performance of the proposed method with existing approach will be presented.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2018 program