Search NASA⌕ Search

Engineering topics

Liang, Feng

Publications and source records attributed to Liang, Feng.

14 records
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Quantifying Epistemic Uncertainty in Binary Classification via Accuracy Gain

ABSTRACT Recently, a surge of interest has been given to quantifying epistemic uncertainty (EU), the reducible portion of uncertainty due to lack of data. We propose a novel EU estimator in the binary classification setting, as the posterior expected value of the empirical gain in accuracy between the current prediction and the optimal prediction. In order to validate the performance of our EU estimator, we introduce an experimental procedure where we take an existing dataset, remove a set of points, and compare the estimated EU with the observed change in accuracy. Through real and simulated data experiments, we demonstrate the effectiveness of our proposed EU estimator.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Improving and Assessing the Quality of Uncertainty Quantification in Deep Learning

Deep learning (DL) models have enjoyed increased attention in recent years because of their powerful predictive capabilities. While many successes have been achieved, standard deep learning methods suffer from a lack of uncertainty quantification (UQ). While the development of methods for producing UQ from DL models is an active area of current research, little attention has been given to the quality of the UQ produced by such methods. In order to deploy DL models to high-consequence applications, high-quality UQ is necessary. This report details the research and development conducted as part of a Laboratory Directed Research and Development (LDRD) project at Sandia National Laboratories. The focus of this project is to develop a framework of methods and metrics for the principled assessment of UQ quality in DL models. This report presents an overview of UQ quality assessment in traditional statistical modeling and describes why this approach is difficult to apply in DL contexts. An assessment on relatively simple simulated data is presented to demonstrate that UQ quality can differ greatly between DL models trained on the same data. A method for simulating image data that can then be used for UQ quality assessment is described. A general method for simulating realistic data for the purpose of assessing a model’s UQ quality is also presented. A Bayesian uncertainty framework for understanding uncertainty and existing metrics is described. Research that came out of collaborations with two university partners are discussed along with a software toolkit that is currently being developed to implement the UQ quality assessment framework as well as serve as a general guide to incorporating UQ into DL applications.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Designing inorganic electrolytes for solid-state Li-ion batteries: A perspective of LGPS and garnet

Solid state Li-ion batteries (SSLBs) are promising technologies for the next generation energy storage devices with higher energy density and enhanced safety. The solid state electrolyte (SSE) is a key component to deliver these desired electrochemical performance characteristics. Here we provide a brief review on the discovery, synthesis, structure, ion-conduction mechanism, and application of LGPS-type and garnet-type lithium ion conductors as two representative SSEs, aiming to extract principles for future design and discovery of favourable solid state Li-ion electrolytes for SSLBs. Recent advances in the strategies to address the challenges with respect to SSLBs are also discussed. Lastly, a perspective on the future research directions of SSLBs is provided.

25 ENERGY STORAGE↗
DOE OSTI2017

Quantifying model structural error: Efficient Bayesian calibration of a regional groundwater flow model using surrogates and a data-driven error model

Groundwater model structural error is ubiquitous, due to simplification and/or misrepresentation of real aquifer systems. During model calibration, the basic hydrogeological parameters may be adjusted to compensate for structural error. This may result in biased predictions when such calibrated models are used to forecast aquifer responses to new forcing. Here, we investigate the impact of model structural error on calibration and prediction of a real-world groundwater flow model, using a Bayesian method with a data-driven error model to explicitly account for model structural error. The error-explicit Bayesian method jointly infers model parameters and structural error and thereby reduces parameter compensation. In this study, Bayesian inference is facilitated using high performance computing and fast surrogate models (based on machine learning techniques) as a substitute for the computationally expensive groundwater model. We demonstrate that with explicit treatment of model structural error, the Bayesian method yields parameter posterior distributions that are substantially different from those derived using classical Bayesian calibration that does not account for model structural error. We also found that the error-explicit Bayesian method gives signficantly more accurate prediction along with reasonable credible intervals. Finally, through variance decomposition, we provide a comprehensive assessment of prediction uncertainty contributed from parameter, model structure, and measurement uncertainty. The results suggest that the error-explicit Bayesian approach provides a solution to real-world modeling applications for which data support the presence of model structural error, yet model deficiency cannot be specifically identified or corrected.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2017

Bayesian calibration of groundwater models with input data uncertainty

Effective water resources management typically relies on numerical models to analyze groundwater flow and solute transport processes. Groundwater models are often subject to input data uncertainty, as some inputs (such as recharge and well pumping rates) are estimated and subject to uncertainty. Current practices of groundwater model calibration often overlook uncertainties in input data; this can lead to biased parameter estimates and compromised predictions. Through a synthetic case study of surface–ground water interaction under changing pumping conditions and land use, we investigate the impacts of uncertain pumping and recharge rates on model calibration and uncertainty analysis. We then present a Bayesian framework of model calibration to handle uncertain input of groundwater models. The framework implements a marginalizing step to account for input data uncertainty when evaluating likelihood. It was found that not accounting for input uncertainty may lead to biased, overconfident parameter estimates because parameters could be over–adjusted to compensate for possible input data errors. Parameter compensation can have deleterious impacts when the calibrated model is used to make forecast under a scenario that is different from calibration conditions. By marginalizing input data uncertainty, the Bayesian calibration approach effectively alleviates parameter compensation and gives more accurate predictions in the synthetic case study. The marginalizing Bayesian method also decomposes prediction uncertainty into uncertainties contributed by parameters, input data, and measurements. Furthermore, the results underscore the need to account for input uncertainty to better inform postmodeling decision making.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2016

Sirtuin1 promotes osteogenic differentiation through downregulation of peroxisome proliferator-activated receptor γ in MC3T3-E1 cells

Osteoporosis is a skeletal disorder characterized by bone loss, resulting in architectural deterioration of the skeleton, decreased bone strength and an increased risk of fragility fractures. Strengthening osteogenesis is an effective way to relieve osteoporosis. Sirtuin1 (Sirt1) is a nicotinamide adenine dinucleotide (NAD{sup +})-dependent deacetylase, which is reported to be involved in improving osteogenesis. Sirt1 targets peroxisome proliferator-activated receptor γ (PPARγ) in the regulation of adipose tissues; however, the molecular mechanism of Sirt1 in osteogenic differentiation is still unknown. PPARγ tends to induce more adipogenic differentiation rather than osteogenic differentiation. Hence, we hypothesized that Sirt1 facilitates osteogenic differentiation through downregulation of PPARγ signaling. Mouse pre-osteoblastic MC3T3-E1 cells were cultured under osteogenic medium. Sirt1 was overexpressed through plasmid transfection. The results showed that high expression of Sirt1 was associated with increased osteogenic differentiation, as indicated by quantitative PCR and Western blot analysis of osteogenic markers, and Von Kossa staining. Sirt1 overexpression also directly and negatively regulated the expression of PPARγ and its downstream molecules. Use of the PPARγ agonist Rosiglitazone, reversed the effects of Sirt1 on osteogenic differentiation. Using constructed luciferase plasmids, we demonstrated a role of Sirt1 in inhibiting PPARγ–induced activity and expression of adipocyte–specific genes, including acetyl-coenzyme A carboxylase (Acc) and fatty acid binding protein 4 (Fabp4). The interaction between Sirt1 and PPARγ was further confirmed using co-immunoprecipitation analysis. Together, these results reveal a novel mechanism for Sirt1 in osteogenic differentiation through downregulation of PPARγ activity. These findings suggest that the Sirt1–PPARγ pathway may represent a potential target for enhancement of osteogenesis and treatment of osteoporosis. - Highlights: • Sirt1 inhibits PPARγ signaling in MC3T3-E1 cells. • PPARγ negatively regulates osteogenic differentiation of MC3T3-E1 cells. • Sirt1 promotes osteogenic differentiation through downregulation of PPARγ.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2015

Reversible switching of ferromagnetism in ZnCuO nanorods by electric field

The reproducible switching of ferromagnetism in ZnCuO nanorods by applying a reversible electric field has been realized. High-resolution transmission electron microscopy images showed a hexagonal wurtzite structure with no detectable trace of secondary phase or precipitation of Cu impurity in the ZnCuO nanorods. The Cu concentrations in the ZnCuO nanorods were tested by energy dispersive spectroscopy and x-ray photoelectron spectroscopy and found to be about 2.7 at. %. The switching mechanism is confirmed in terms of the formation and rupture of conductive filaments, with oxygen vacancies (V{sub O}) localized mainly on surface of the ZnCuO nanorods. Subsequently, the variation of V{sub O} concentration during the resistive switching process modulates the ferromagnetism of the ZnCuO nanorods. The saturation magnetization at low resistance state is apparently 6.4 times larger than that at high resistance state for an Au/ZnCuO/ITO structure. An indirect double-exchange model has been used to explain the ferromagnetism in ZnCuO nanorods.

77 NANOSCIENCE AND NANOTECHNOLOGY↗
DOE OSTI2014

Measurement of anode surface temperature in carbon nanomaterial production by arc discharge method

Highlights: • We measured the temperature of anode surface by two-color pyrometry combined with a high speed camera successfully. • Growth temperature of pyrolytic graphite, MWNTs, and nano-graphite particles were in ranges of 2400–2600 K, 2600–2700 K, and 2700–3500 K, respectively. • High temperature contributes to form thermodynamically unstable material. - Abstract: Nano-graphite particles, multi-wall carbon nanotube (MWNT), and pyrolytic graphite were prepared at different positions of the anode surface in an arc discharge. Graphite electrodes were employed for the arc discharge under helium environment at atmospheric pressure. Nano-sized carbon products were characterized by scanning electron microscopy and transmission electron microscopy. During the arc discharge, two-color pyrometry combined with a high-speed camera was conducted to measure the temperature distribution of the anode surface. The growth temperature of pyrolytic graphite, MWNT, and nano-graphite particles were in the ranges of 2400–2600 K, 2600–2700 K, and 2700–3500 K, respectively. The local temperature of anode surface is a critical parameter to determine the products with different morphologies. The formation mechanism of these carbon nanomaterials is suggested based on the local temperature of anode surface and their thermodynamic stability.

36 MATERIALS SCIENCE↗
DOE OSTI2014

Physical origin of nanograting formation on fused silica with femtosecond pulses

We present a comprehensive analysis of physical evolution of nanograting formation based on an experiment performed with femtosecond pulses focused under moderate focusing conditions and where pulse energy is slowly increased as the focused beam is moved along the sample surface. The results demonstrate that nanograting inscription is initiated at the location of the maximum plasma density and evolves through local intensity side lobes, whose locations are self-regulated in a closed feedback loop, in agreement with the plasmonic model.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2014

High sensitivity and high Q-factor nanoslotted parallel quadrabeam photonic crystal cavity for real-time and label-free sensing

We experimentally demonstrate a label-free sensor based on nanoslotted parallel quadrabeam photonic crystal cavity (NPQC). The NPQC possesses both high sensitivity and high Q-factor. We achieved sensitivity (S) of 451 nm/refractive index unit and Q-factor >7000 in water at telecom wavelength range, featuring a sensor figure of merit >2000, an order of magnitude improvement over the previous photonic crystal sensors. In addition, we measured the streptavidin-biotin binding affinity and detected 10 ag/mL concentrated streptavidin in the phosphate buffered saline solution.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗