Shandian Zhe

About

Shandian Zhe is an Associate Professor in the School of Computing at the University of Utah. He obtained his Ph.D. from Purdue University in 2017. His research focuses on probabilistic learning, including probabilistic graphical models, Bayesian deep neural networks, operator learning, and kernel methods, with applications in biomedical data analysis and online advertising.

Education

• Ph.D. Computer Science, Purdue University, 2017

Selected publications

• Kronecker-Structured Nonparametric Spatiotemporal Point Processes
• Complexity-Aware Deep Symbolic Regression with Robust Risk-Seeking Policy Gradients
• Tensor Gaussian Processes: Efficient Solvers for Nonlinear PDEs
• ElastoGen: 4D Generative Elastodynamics
• Kernel Neural Operators (KNOs) for Scalable, Memory-efficient, Geometrically-flexible Operator Learning
• StFT: Spatio-temporal Fourier Transformer for Long-term Dynamics Prediction
• A Streaming Sparse Cholesky Method for Derivative-Informed Gaussian Process Surrogates Within Digital Twin Applications
• Efficiently Training SciML Models with Derivative-Informed Training Data Using Order Truncated Imaginary Numbers
• Polynomial-Augmented Neural Networks (PANNS) With Weak Orthogonality Constraints For Enhanced Function and PDE Approximations
• Using Residual Analysis to Characterize and Control the Impact of Noisy Data on Stress Intensity Factor Models from Machine Learning