Search NASA⌕ Search

Engineering topics

Guo, Ling

Publications and source records attributed to Guo, Ling.

Uncertainty quantification in scientific machine learning: Methods, metrics, and comparisons

Neural networks (NNs) are currently changing the computational paradigm on how to combine data with mathematical laws in physics and engineering in a profound way, tackling challenging inverse and ill-posed problems not solvable with traditional methods. However, quantifying errors and uncertainties in NN-based inference is more complicated than in traditional methods. This is because in addition to aleatoric uncertainty associated with noisy data, there is also uncertainty due to limited data, but also due to NN hyperparameters, overparametrization, optimization and sampling errors as well as model misspecification. Although there are some recent works on uncertainty quantification (UQ) in NNs, there is no systematic investigation of suitable methods towards quantifying the total uncertainty effectively and efficiently even for function approximation, and there is even less work on solving partial differential equations and learning operator mappings between infinite-dimensional function spaces using NNs. In this work, we present a comprehensive framework that includes uncertainty modeling, new and existing solution methods, as well as evaluation metrics and post-hoc improvement approaches. Further, to demonstrate the applicability and reliability of our framework, we present an extensive comparative study in which various methods are tested on prototype problems, including problems with mixed input-output data, and stochastic problems in high dimensions. In the Appendix, we include a comprehensive description of all the UQ methods employed. Further, to help facilitate the deployment of UQ in Scientific Machine Learning research and practice, we present and develop in [1] an open-source Python library (github.com/Crunch-UQ4MI/neuraluq), termed NeuralUQ, that is accompanied by an educational tutorial and additional computational experiments.

11 physics-informed neural networks↗

Learning in Modal Space: Solving Time-Dependent Stochastic PDEs Using Physics-Informed Neural Networks

One of the open problems in scientific computing is the long-time integration of nonlinear stochastic partial differential equations (SPDEs), especially with arbitrary initial data. We address this problem by taking advantage of recent advances in scientific machine learning and the spectral dynamically orthogonal (DO) and borthogonal (BO) methods for representing stochastic processes. The recently introduced DO/BO methods reduce the SPDE to solving a system of deterministic PDEs and a system of stochastic ordinary differential equations. Specifically, we propose two new physics-informed neural networks (PINNs) for solving time-dependent SPDEs, namely the neural network (NN)-DO/BO methods. The proposed methods incorporate the DO/BO constraints into the loss function (along with the modal decomposition of the SPDE) with an implicit form instead of generating explicit expressions for the temporal derivatives of the DO/BO modes. Hence, the NN-DO/BO methods can overcome some of the drawbacks of the original DO/BO methods. For example, we do not need the assumption that the covariance matrix of the random coefficients is invertible as in the original DO method, and we can remove the assumption of no eigenvalue crossing as in the original BO method. Moreover, the NN-DO/BO methods can be used to solve time-dependent stochastic inverse problems with the same formulation and same computational complexity as for forward problems. Furthermore, we demonstrate the capability of the proposed methods via several numerical examples, namely: (1) A linear stochastic advection equation with deterministic initial condition: we obtain good results with the proposed methods, while the original DO/BO methods cannot be applied directly in this case. (2) Long-time integration of the stochastic Burgers' equation: we show the good performance of NN-DO/BO methods, especially the effectiveness of the NN-BO approach for such problems with many eigenvalue crossings during the whole time evolution, while the original BO method fails. (3) Nonlinear reaction diffusion equation: we consider both the forward problem and the inverse problems, including very noisy initial point values, to investigate the flexibility of the NN-DO/BO methods in handling inverse and mixed type problems. Taken together, these simulation results demonstrate that the NN-DO/BO methods can be employed to effectively quantify uncertainty propagation in a wide range of physical problems, but future work should address the efficiency issue of PINNs for forward problems.

97 MATHEMATICS AND COMPUTING↗

The Prognostic Value of Plasma Epstein-Barr Viral DNA and Tumor Response to Neoadjuvant Chemotherapy in Advanced-Stage Nasopharyngeal Carcinoma

Purpose: To explore the prognostic value of the plasma load of Epstein-Barr viral (EBV) DNA and the tumor response to neoadjuvant chemotherapy (NACT) in advanced-stage nasopharyngeal carcinoma (NPC). Patients and Methods: In all, 185 consecutive patients with stage III to IVb NPC treated with NACT followed by concurrent chemoradiation therapy (CCRT) were prospectively enrolled. The primary endpoint was progression-free survival (PFS), and the secondary endpoints included locoregional relapse–free survival (LRFS) and distant metastasis–free survival (DMFS). Results: EBV DNA was detected in 165 (89%) patients before treatment but was undetectable in 127 (69%) patients after NACT. Detectable EBV DNA levels after NACT were correlated with poor prognosis (3-year PFS 71.8% vs 85.2%, P=.008 and 3-year DMFS 82.5% vs 92.3%, P=.013). An unsatisfactory tumor response (stable disease or disease progression) after NACT was also correlated with poor clinical outcome (3-year PFS 71.1% vs 85.9%, P=.005 and 3-year LRFS 82.7% vs 93.5%, P=.012). Multivariate analysis showed that the EBV DNA level after NACT (hazard ratio [HR] 2.31, 95% CI 1.18-4.54, P=.015) and the tumor response to NACT (HR 2.84, 95% CI 1.42-5.67, P=.003) were both significant prognostic factors for PFS. Multivariate analysis also showed that EBV DNA after NACT was the only significant predictor of DMFS (HR 2.99, 95% CI 1.25-7.15, P=.014) and that tumor response to NACT was the only significant predictor of LRFS (HR 3.31, 95% CI 1.21-9.07, P=.020). Conclusion: Detectable EBV DNA levels and an unsatisfactory tumor response (stable disease or disease progression) after NACT serve as predictors of poor prognosis for patients with advanced-stage NPC. These findings will facilitate further risk stratification, early treatment modification, or both before CCRT.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Cadmium induces vascular permeability via activation of the p38 MAPK pathway

Highlights: • Low-dose cadmium (Cd) induces vascular hyper-permeability. • p38 MAPK mediates Cd-induced disruption of endothelial cell barrier function. • SB203850 inhibits Cd-induced membrane dissociation of VE-cadherin and β-catenin. • SB203850 reduces Cd-induced expression and secretion of TNF-α. - Abstract: The vasculature of various organs is a targeted by the environmental toxin, cadmium (Cd). However, mechanisms leading to pathological conditions are poorly understood. In the present study, we examined the effect of cadmium chloride (CdCl{sub 2}) on human umbilical vein endothelial cells (HUVECs). At 4 μM, CdCl{sub 2} induced a hyper-permeability defect in HUVECs, but not the inhibition of cell growth up to 24 h. This effect of CdCl{sub 2} was dependent on the activation of the p38 mitogen-activated protein kinase (MAPK) pathway. The p38 MAPK inhibitor SB203850 suppressed the CdCl{sub 2}-induced alteration in trans-endothelial electrical resistance in HUVEC monolayers, a model measurement of vascular endothelial barrier integrity. SB203850 also inhibited the Cd-induced membrane dissociation of vascular endothelial (VE) cadherin and β-catenin, the important components of the adherens junctional complex. In addition, SB203850 reduces the Cd-induced expression and secretion of tumor necrosis factor α (TNF-α). Taken together, our findings suggest that Cd induces vascular hyper-permeability and disruption of endothelial barrier integrity through stimulation of p38 MAPK signaling.

60 APPLIED LIFE SCIENCES↗

Theoretical simulation of carrier capture and relaxation rates in quantum-dot semiconductor optical amplifiers

Based on Auger scattering mechanism, carrier-carrier scattering dynamics between the two-dimensional carrier reservoir (also called wetting layer, i.e., WL) and the confined quantum dot ground and first excited state in quantum-dot semiconductor optical amplifiers (QD-SOAs) are investigated theoretically in this paper. The scattering rates for independent electron and hole densities are calculated. The results show an ultra-fast carrier capture (relaxation) rate up to 1 ps{sup −1}, and there is a complex dependence of the Coulomb scattering rates on the WL electron and hole densities. In addition, due to the different effective mass and the level distribution, the scattering rates for electron and hole are very different. Finally, in order to provide a direction to control (increase or decrease) the input current in realistic QD-SOA systems, a simple method is proposed to determine the trends of the carrier recovery rates with the WL carrier densities in the vicinity of the steady-state.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗