Papers written by doctoral students in the H. Milton Stewart School of Industrial and Systems Engineering (ISyE) – fifth-year Jialei Chen, fifth-year Ana María Estrada Gómez, and third-year Henry Shaowu Yuchi – have been selected for awards by the Sections on Physical and Engineering Sciences (SPES) and Quality and Productivity of the American Statistical Association. ISyE composes three of the five total award winners.
Chen’s paper, “APIK: A Physics-Informed Kriging Model with Partial Differential Equations,” focuses on the challenge of applying state-of-the-art machine learning methods in real-world engineering applications when available measurement data is scarce. He advances a new learning model that leverages both data and auxiliary partial differential equations, applying the proposed method to two real-world applications on flow dynamics and thermal processes. His co-authors include Ph.D. student Zhehui Chen, Harold E. Smalley Professor Chuck Zhang, and Coca-Cola Chair in Engineering Statistics and Professor Jeff Wu.
Chen’s research focuses on engineering-driven learning methodologies, and data-driven modeling for complex engineering and manufacturing systems. He is advised by A. Russell Chandler III Professor Roshan Joseph and Zhang.
Estrada’s paper, “An Adaptive Sampling Strategy for Online Monitoring and Diagnosis of High-Dimensional Streaming Data,” presents a new adaptive sampling strategy for online monitoring and diagnosis under resource constraints. The proposed methodology integrates two novel ideas: (1) the recursive projection of the high-dimensional streaming data onto a low-dimensional subspace to capture the spatio-temporal structure of the data while performing missing data imputation; and (2) the development of an adaptive sampling scheme, balancing exploration and exploitation, to decide where to collect data at each acquisition time. The adaptive sampling strategy was used to improve soil-water temperature monitoring in a dryland agricultural field to ensure the crops' quality. Her co-authors include Ph.D. student Dan Li and Fouts Family Early Career Professor and Associate Professor Kamran Paynabar, who is her advisor.
Estrada’s research interests revolve around developing efficient methodologies and algorithms for modeling, monitoring, and diagnosing sensing systems with high-dimensional data using statistical and machine learning tools.
Yuchi’s paper, “Computer Experiments with Multiple Mesh Density Variables,” proposes new modeling methods for deterministic computer experiments with multiple tuning parameters. The work uses a non-stationary Gaussian process model to bring together simulations of different mesh densities in finite element analysis and to improve the overall prediction performance. Yuchi’s co-authors are Professors Joseph and Wu.
Yuchi is advised by Wu and Harold R. and Mary Anne Nash Early Career Professor and Associate Professor Yao Xie.His research generally is on the topic of multi-fidelity and examines how to use statistical methods to model practical applications.
The three students received travel awards that will support their attendance at the 2021 Joint Statistical Meetings in Seattle, Washington, in early August.