Efficient Bayesian inference using adversarial machine learning and low-complexity surrogate models
| Autor: | Jonggeol Na, Ji Hyun Bak, Nikolaos V. Sahinidis |
|---|---|
| Rok vydání: | 2021 |
| Předmět: |
Optimization problem
Estimation theory business.industry Computer science 020209 energy General Chemical Engineering Inference 02 engineering and technology Adversarial machine learning Bayesian inference Machine learning computer.software_genre Computer Science Applications Hybrid Monte Carlo 020401 chemical engineering 0202 electrical engineering electronic engineering information engineering Artificial intelligence 0204 chemical engineering business Likelihood function computer Parametric statistics |
| Zdroj: | Computers & Chemical Engineering. 151:107322 |
| ISSN: | 0098-1354 |
| DOI: | 10.1016/j.compchemeng.2021.107322 |
| Popis: | Bayesian inference is a key method for estimating parametric uncertainty from data. However, most Bayesian inference methods require the explicit likelihood function or many samples, both of which are unrealistic to provide for complex first-principles-based models. Here, we propose a novel Bayesian inference methodology for estimating uncertain parameters of computationally intensive first-principles-based models. Our approach exploits both low-complexity surrogate models and variational inference with arbitrarily expressive inference models. The proposed methodology indirectly predicts output responses and casts Bayesian inference as an optimization problem. We demonstrate its performance via synthetic problems, computational fluid dynamics, and kinetic Monte Carlo simulation to verify its applicability. This fast and reliable methodology enables us to capture multimodality and the shape of complicated posterior distributions with the quality of state-of-the-art Hamiltonian Monte Carlo methods but with much lower computation cost. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full Text from ScienceDirect