ASA Connect

Dear ASA Community,

We are pleased to continue the Spring 2026 edition of the TIES Environmetrics Webinar Series with our upcoming talk on February 27 at 11:00 AM Eastern Time.

You can access the webinar and register via our website (if you have previously registered with us, no additional registration is needed):

Won Chang, Seoul National University

Fast Computer Model Calibration using Annealed and Transformed Variational Inference

Computer models play a crucial role in numerous scientific and engineering domains. To ensure the accuracy of simulations, it is essential to properly calibrate the input parameters of these models through statistical inference. While Bayesian inference is the standard approach for this task, employing Markov chain Monte Carlo methods often encounters computational hurdles due to the costly evaluation of likelihood functions and slow mixing rates. Although variational inference (VI) can be a fast alternative to traditional Bayesian approaches, VI has limited applicability due to boundary issues and local optima problems. To address these challenges, we propose flexible VI methods based on deep generative models that do not require parametric assumptions on the variational distribution. We embed a surjective transformation in our framework to avoid posterior truncation at the boundary. Additionally, we provide theoretical conditions that guarantee the success of the algorithm. Furthermore, our temperature annealing scheme and fine-tuning can prevent being trapped in local optima through a series of intermediate posteriors and weight adjustment. We apply our method to infectious disease models and a geophysical model, illustrating that the proposed method can provide fast and accurate inference compared to its competitors.

Won Chang is an Associate Professor in the Department of Statistics at Seoul National University. He received his Ph.D. in Statistics from Pennsylvania State University and previously served on the faculty at the University of Cincinnati before joining SNU in 2024. His research focuses on statistical methodology, uncertainty quantification, and data-driven modeling, with particular emphasis on computer model calibration, Bayesian statistics, and spatial data analysis. More recently, his work has centered on utilizing deep learning models and deep generative models for Bayesian inference and computer model calibration, aiming to scale rigorous statistical inference to complex, high-dimensional problems. He has published extensively on methods for complex environmental and biological data, bridging theoretical development with applications in atmospheric science, hydrology, genetics, and climate modeling.

We hope to see you there!