SPES and Q&P Student Paper Award

We propose a novel network generative model extended from the standard stochastic block model by concurrently utilizing observed node-level information and accounting for network enabled nodal heterogeneity. The proposed model is so called covariate-assisted latent factor stochastic block model (CALF-SBM), inference of which is done in a fully Bayesian framework. The primary application of CALF-SBM in the present research is focused on community detection, where a model-selection-based method is employed to estimate the number of communities if it is unknown. To assess the performance of CALF-SBM, an extensive simulation study is carried out, including comparisons with multiple classical and modern network clustering algorithms. Lastly, the paper presents two real data applications respectively based on an extremely new network data demonstrating collaborative relationships of the otolaryngologists in the United States and a traditional aviation network data collecting information about direct flights among major airports in the United States and Canada. 

Surrogate modeling is essential for analyzing expensive computer experiments but often suffers from the curse of dimensionality, limiting its scalability and interpretability. To address these challenges, we propose a novel composite framework that combines the uncertainty quantification of Gaussian process models with the scalability of deep neural networks (DNNs). A key innovation of our approach is the integration of basis expansion with a difference penalty, providing an efficient and interpretable solution for network structure selection. By drawing an analogy between neuron selection in DNNs and knot selection in splines, we extend statistical frameworks to deep learning, enabling automated and computationally efficient structure tuning through an Expectation Conditional Maximization (ECM)-based algorithm. Our penalized framework offers unique flexibility by decoupling penalty order from spline basis degree, enhancing model design. Theoretical analyses reveal that the proposed method achieves superior convergence rates and effectively circumvents the curse of dimensionality, making it suitable for nonlinear regression with high-dimensional inputs. Applications to surrogate modeling demonstrate the framework's versatility and robustness. This work establishes a principled foundation for scalable, interpretable, and theoretically grounded DNN-based surrogate modeling techniques. 

The configurations of instruments fielded on an experiment affect the amount of information captured and the quality of subsequent inference. We investigate the problem of optimizing plasma x-ray radiation detectors in a magneto-inertial fusion experiment at Sandia National Laboratories. It is impossible to directly measure properties such as the temperature of the thermonuclear fusion plasma produced in these experiments because of the extreme environment and destructive nature of the experiment. Among other diagnostics, several detectors are placed with significant standoff from the fusion target to capture the x-rays emitted by the fusion plasma, which can be used to infer some of its properties. To optimize the configuration of these detectors, a high-fidelity model (HFM) is used for simulating outputs and a low-fidelity model (LFM) is used for inference. We develop methods based on A- and L-optimality criteria that are efficient to compute while explicitly accounting for the discrepancy between the HFM and the LFM. The method allows us to find detector configurations that perform similarly to or better than the configuration obtained using an existing sampling-based optimization method while decreasing computational time by a factor of 50. 

One common approach to the statistical analysis of spatially correlated data relies on defining a correlation structure based solely on the physical distance between the locations of observed values. However, some data have a complex spatial structure that cannot be adequately described with the physical distance alone. In this line of research, the spatial failure-time data of focus contains information on GPUs that are linked through a series of wired connections, where it is expected that the failure-times of GPUs with few connections between them will be highly correlated. The proposed lifetime regression model includes random effects capturing the dependency due to physical location as well as random effects explaining the dependency due to the number of logical connections between GPUs. The analysis of this GPU dataset serves as an example of models with multiple spatial random effects and the ideas presented can be extended to other applications with complex spatial structures. A Bayesian modeling scheme is recommended for this class of analyses. The examples in this presentation use the software package, Stan, to produce Markov chain Monte Carlo draws for parameter estimation. This modeling effort is validated through simulation which demonstrates the accuracy of statistical inference. We also apply the developed framework to the large-scale Titan GPU failure-time data. 

Composite materials have long been used in various industries due to their superior properties such as high strength, lightweight, and corrosive resistance. Bonded composite joints are finding increasing applications, as they provide extensive structural benefits and design flexibility. On the other hand, the failure mechanism of composite adhesive joints is not fully understood. A model that bridges manufacturing parameters and final quality measures is highly desired for the design and optimization of the manufacturing process of composite adhesive joints. In this study, a novel framework of Physics-Informed Neural Ordinary Differential Equation (ODE) with Heterogeneous Control Input (PINOHI) is proposed, which links the heterogeneous manufacturing parameters to the final bonding quality of composite joints. The proposed model structure is heavily motivated by engineering knowledge, incorporating a calibrated mathematical physics model into the Neural ODE framework, which can significantly reduce the number of data samples required from costly experiments while maintaining high prediction accuracy. The proposed PINOHI model is implemented in the quality prediction of composite adhesive joints bonding problem. A set of experiments and associated data analytics are conducted to demonstrate the superior property of the PINOHI model by using both the leave-one-batch-out cross-validation and sensitivity analysis.