Research

My current research interest is in System Informatics and Control, with an emphasis on developing AI/ML techniques to fuse domain-specific knowledge (such as data structural information, physical laws, and engineering knowledge) and in-situ sensing data, to enhance the modeling and decision-making for complex systems, which interfaces high-dimensional statistics, optimization research, machine learning, and computational science:

Robust machine learning enabled modeling and monitoring for high dimensional (HD) data streams, with application in process monitoring of advanced manufactring processes.
Large scale optimziation with its applciations in machine learning problems:
- Learning based optimal design of complex engineering systems
- Learning based synthetic data generation and evaluation
Physics informed machine learning for complex systems

I am also interested in the analysis of stochastic systems using tools from queuing theory.

I have been working on various research projects funded by the National Science Foundation, Apple, Boeing Company, Samsung and OG Technologies.

Here you can find a brief overview of my research from my talk at UW-Madison [slides].

Publications (organized by topic):

Robust learning with applications in anomaly detection

Inverse problems, Generative models, Matrix/tensor decomposition

Develop robust, scalable, and computationally efficient machine learning algorithms for modeling and analysis of HD data with complex structure (tensor structure, complex spatiotemporal structure, etc.), with applications in manufacturing anomaly detection, root cause diagnosis and so on.

RGI: robust GAN-inversion for mask-free image inpainting and unsupervised pixel-wise anomaly detection.

The Eleventh International Conference on Learning Representations (ICLR 2023) (Presentation)

Mou, S., Gu, X., Cao, M., Bai, H., Huang, P., Shan, J., & Shi, J. 2023.

2023 IISE QCRE Best Track Paper Award Winner

Compressed Smooth Sparse Decomposition.

INFORMS Journal on Data Science https://doi.org/10.1287/ijds.2022.0023

Mou, S., Shi, J.* 2022.

2022 IISE DAIS Best Track Paper Award Finalist

PAEDID: Patch Autoencoder based Deep Image Decomposition.

IISE Transactions https://doi.org/10.1080/24725854.2022.2163435.

Mou, S., Cao, M., Bai, H., Huang, P., Shi, J.*, Shan, J. 2022.

Additive Tensor Decomposition Considering Structural Data Information.

IEEE Transactions on Automation Science and Engineering https://doi.org/10.1109/TASE.2021.3096964

Mou, S., Wang, A., Zhang, C., & Shi, J.* 2021.

2021 IISE DAIS Best Student Paper Award Finalist

Design and optimization of complex physical systems with applications in advanced manufacturing

Scientific machine learning, PDE constrained optimization, numerical methods in PDE and Tensor Decomposition.(Note: † indicates a student I mentored.)

Develop physics-informed machine learning/PDE-constrained optimization algorithms to synergize physical knowledge and process data with applications in process control, quality engineering, and design optimization of multistage manufacturing systems.