Session: DAC-11-01: Design of Engineering Materials and Structures

Paper Number: 114414

114414 - Mitigating the Effects of Source-Dependent Bias and Noise on Multi-Source Bayesian Optimization: Application to Materials Design 

Bayesian optimization (BO) is a sequential optimization strategy that is increasingly employed in a wide range of areas including materials design and drug discovery. In real world applications, acquiring high-fidelity (HF) data is the major cost component of BO, since most of the problems demand the use of expensive HF simulations. To alleviate this bottleneck, multi-fidelity (MF) methods are proposed to forgo the sole reliance on the expensive HF data and reduce the sampling costs by querying inexpensive low-fidelity (LF) sources whose data are correlated with HF samples. Existing multi-fidelity BO (MFBO) methods operate under the following two assumptions: (1) Leveraging global (rather than local) correlation between HF and LF sources, and (2) Associating all the data sources with the same noise process. These assumptions dramatically reduce the performance of MFBO when LF sources are only locally correlated with the HF source or when the noise variance varies across the data sources. To dispense with these incorrect assumptions, we propose an MF emulation method that learns a source-dependent noise process and also enables BO to leverage highly biased LF sources which are only locally correlated with the HF source. We illustrate the performance of our method through analytical examples and engineering problems on materials design.

Presenting Author: Zahra Zanjani Foumani University of California Irvine

Presenting Author Biography: My name is Sanaz. I'm second year PhD student of Mechanical and Aerospace engineering at University of California Irvine. I got my bachelors' degree on Industrial Engineering from University of Tehran.

Authors:

Zahra Zanjani Foumani University of California Irvine
Amin Yousefpour University of California Irvine
Mehdi Shishehbor University of California Irvine
Ramin Bostanabad University of California Irvine