Bayesian inference of elastic properties with resonant ultrasound spectroscopy
Not provided.
SEARCH · Engineering Papers
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Not provided.
Not provided.
Uncertainty quantification (UQ) plays a pivotal role in scientific machine learning, especially when surrogate models are used to approximate complex systems. Although multilayer perceptions (MLPs) are commonly employed as surrogates, they often suffer from overfitting due to their large number of parameters. Kolmogorov-Arnold networks (KANs) offer an alternative solution with fewer parameters. However, gradient-based inference methods, such as Hamiltonian Monte Carlo (HMC), may result in computational inefficiency when applied to KANs, especially for large-scale datasets, due to the high cost of back-propagation. To address these challenges, we propose a novel approach, combining the dropout Tikhonov ensemble Kalman inversion (DTEKI) with Chebyshev KANs. This gradient-free method effectively mitigates overfitting and enhances numerical stability. In addition, we incorporate the active subspace method to reduce the parameter-space dimensionality, allowing us to improve the accuracy of predictions and obtain more reliable uncertainty estimates. Extensive experiments demonstrate the efficacy of our approach in various test cases, including scenarios with large datasets and high noise levels. Our results show that the new method achieves comparable or better accuracy, much higher efficiency as well as stability compared to HMC, in addition to scalability. Moreover, by leveraging the low-dimensional parameter subspace, our method preserves prediction accuracy while substantially reducing further the computational cost.
Many key building design policies are made using sophisticated computer simulations such as EnergyPlus (E+), the DOE flagship whole-building energy simulation engine. E+ and other sophisticated computer simulations have several major problems. The two main issues are 1) gaps between the simulation model and the actual structure, and 2) limitations of the modeling engine's capabilities. Currently, these problems are addressed by having an engineer manually calibrate simulation parameters to real world data or using algorithmic optimization methods to adjust the building parameters. However, some simulations engines, like E+, are computationally expensive, which makes repeatedly evaluating the simulation engine costly. This work explores addressing this issue by automatically discovering the simulation's internal input and output dependencies from ~20 Gigabytes of E+ simulation data, future extensions will use ~200 Terabytes of E+ simulation data. The model is validated by inferring building parameters for E+ simulations with ground truth building parameters. Our results indicate that the model accurately represents parameter means with some deviation from the means, but does not support inferring parameter values that exist on the distribution's tail.
Explore the source record for details and available documents.
Neuromorphic systems allow for extremely efficient hardware implementations for neural networks (NNs). In recent years, several algorithms have been presented to train spiking NNs (SNNs) for neuromorphic hardware. However, SNNs often provide lower accuracy than their artificial NNs (ANNs) counterparts or require computationally expensive and slow training/inference methods. To close this gap, designers typically rely on reconfiguring SNNs through adjustments in the neuron/synapse model or training algorithm itself. Nevertheless, these steps incur significant design time, while still lacking the desired improvement in terms of training/inference times (latency). Designing SNNs that can mimic the accuracy of ANNs with reasonable training times is an exigent challenge in neuromorphic computing. In this work, we present an alternative approach that looks at such designs as an optimization problem rather than algorithm or architecture redesign. We develop a versatile multiobjective hyperparameter optimization (HPO) for automatically tuning HPs of two state-of-the-art SNN training algorithms, SLAYER and HYBRID. We emphasize that, to the best of our knowledge, this is the first work trying to improve SNNs’ computational efficiency, accuracy, and training time using an efficient HPO. We demonstrate significant performance improvements for SNNs on several datasets without the need to redesign or invent new training algorithms/architectures. Our approach results in more accurate networks with lower latency and, in turn, higher energy efficiency than previous implementations. In particular, we demonstrate improvement in accuracy and more than 5× reduction in the training/inference time for the SLAYER algorithm on the DVS Gesture dataset. In the case of HYBRID, we demonstrate 30% reduction in timesteps while surpassing the accuracy of the state-of-the-art networks on CIFAR10. Further, our analysis suggests that even a seemingly minor change in HPs could change the accuracy by 5 - 6×.
System codes for simulation of safety performance of nuclear plants may contain parameters whose values are not known very accurately. New information from tests or operating experience is incorporated into safety codes by a process known as calibration, which reduces uncertainty in the output of the code and thereby improves its support for decision-making. The work reported here implements several improvements on classic calibration techniques afforded by modern analysis techniques. The key innovation has come from development of code surrogate model (or code emulator) construction and prediction algorithms. Use of a fast emulator makes the calibration processes used here with Markov Chain Monte Carlo (MCMC) sampling feasible. This study uses Gaussian Process (GP) based emulators, which have been used previously to emulate computer codes in the nuclear field. The present work describes the formulation of an emulator that incorporates GPs into a factor analysis-type or pattern recognition-type model. This “function factorization” Gaussian Process (FFGP) model allows overcoming limitations present in standard GP emulators, thereby improving both accuracy and speed of the emulator-based calibration process. Calibration of a friction-factor example using a Method of Manufactured Solution is performed to illustrate key properties of the FFGP based process.
Bayesloc is a statistical model of the multiple seismic location system, including event hypocenters, corrections to model-based travel time predictions, assessments precision for measurement phase arrival times, and phase lavels which indicate phase ray path.
BOOTS is a package for optimal selection of candidate operating points in large-scale manufacturing applications. It is designed to optimally select operating points to maximize predicted values of product quality and resource efficiency according to a data-driven model. The package is based on the use of a multi-input multi-output (MIMO) Gaussian process model to describe the relationships between inputs (operating points) and outputs (product quality and resource efficiency). This package does not provide any site- or process-specific information.
Not Available
We introduce here an approach based on the Givens representation for posterior inference in statistical models with orthogonal matrix parameters, such as factor models and probabilistic principal component analysis (PPCA). We show how the Givens representation can be used to develop practical methods for transforming densities over the Stiefel manifold into densities over subsets of Euclidean space. We demonstrate how to deal with issues arising from the topology of the Stiefel manifold and how to inexpensively compute the change-of-measure terms. We introduce an auxiliary parameter approach that limits the impact of topological issues. We provide both analysis of our methods and numerical examples demonstrating the effectiveness of the approach. We also discuss how our Givens representation can be used to define general classes of distributions over the space of orthogonal matrices. We then give demonstrations on several examples showing how the Givens approach performs in practice in comparison with other methods.
Explore the source record for details and available documents.
From nuclear safeguards to homeland security applications, the need for the better modeling of nuclear interactions has grown over the past decades. Current Monte Carlo radiation transport codes compute average quantities with great accuracy and performance; however, performance and averaging come at the price of limited interaction-by-interaction modeling. These codes often lack the capability of modeling interactions exactly: for a given collision, energy is not conserved, energies of emitted particles are uncorrelated, and multiplicities of prompt fission neutrons and photons are uncorrelated. Many modern applications require more exclusive quantities than averages, such as the fluctuations in certain observables (e.g., the neutron multiplicity) and correlations between neutrons and photons. In an effort to meet this need, the radiation transport Monte Carlo code TRIPOLI-4® was modified to provide a specific mode that models nuclear interactions in a full analog way, replicating as much as possible the underlying physical process. Furthermore, the computational model FREYA (Fission Reaction Event Yield Algorithm) was coupled with TRIPOLI-4 to model complete fission events. As a result, FREYA automatically includes fluctuations as well as correlations resulting from conservation of energy and momentum.