|
Activity Number:
|
571
- Statistics for Computer Experiments: Collaboration Between Industry and Academia
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Wednesday, August 2, 2017 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Physical and Engineering Sciences
|
|
Abstract #323457
|
|
|
Title:
|
Exact Knowledge Gradient for Stochastic Computer Model Assisted Optimal Decision Making
|
|
Author(s):
|
QIONG ZHANG* and Youngdeok Hwang
|
|
Companies:
|
Virginia Commonwealth University and IBM Thomas J. Watson Research Center
|
|
Keywords:
|
Knowledge gradient ;
Model-based optimization ;
Gaussian process ;
Computer Experiments
|
|
Abstract:
|
Optimization using stochastic computer experiments is commonplace in engineering and industry. The article addresses the problem of optimization when the input space of stochastic computer model is continuous, whereas the decision space in real problem is restricted to be discrete. We propose new exact knowledge gradient based Bayesian sequential selection methods for optimal decision making. We demonstrate that our proposed methods are competitive compared to existing model-based optimization methods.
|
Authors who are presenting talks have a * after their name.
