SEARCH · Engineering Papers
Results for “sampling algorithm”
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.
Lecture 1: Multilevel-Multifidelity with Monte Carlo Sampling; Algorithms and deployment experience.
Abstract not provided.
A parallel evolutionary multiple-try metropolis Markov chain Monte Carlo algorithm for sampling spatial partitions
We develop an Evolutionary Markov Chain Monte Carlo (EMCMC) algorithm for sampling spatial partitions that lie within a large, complex, and constrained spatial state space. Our algorithm combines the advantages of evolutionary algorithms (EAs) as optimization heuristics for state space traversal and the theoretical convergence properties of Markov Chain Monte Carlo algorithms for sampling from unknown distributions. Local optimality information that is identified via a directed search by our optimization heuristic is used to adaptively update a Markov chain in a promising direction within the framework of a Multiple-Try Metropolis Markov Chain model that incorporates a generalized Metropolis-Hastings ratio. We further expand the reach of our EMCMC algorithm by harnessing the computational power afforded by massively parallel computing architecture through the integration of a parallel EA framework that guides Markov chains running in parallel.
Source term estimation using multiple xenon isotopes in atmospheric samples
Not Available
Final Report: Sampling-Based Algorithms for Estimating Structure in Big Data.
The purpose of this project was to develop sampling-based algorithms to discover hidden struc- ture in massive data sets. Inferring structure in large data sets is an increasingly common task in many critical national security applications. These data sets come from myriad sources, such as network traffic, sensor data, and data generated by large-scale simulations. They are often so large that traditional data mining techniques are time consuming or even infeasible. To address this problem, we focus on a class of algorithms that do not compute an exact answer, but instead use sampling to compute an approximate answer using fewer resources. The particular class of algorithms that we focus on are streaming algorithms , so called because they are designed to handle high-throughput streams of data. Streaming algorithms have only a small amount of working storage - much less than the size of the full data stream - so they must necessarily use sampling to approximate the correct answer. We present two results: * A streaming algorithm called HyperHeadTail , that estimates the degree distribution of a graph (i.e., the distribution of the number of connections for each node in a network). The degree distribution is a fundamental graph property, but prior work on estimating the degree distribution in a streaming setting was impractical for many real-world application. We improve upon prior work by developing an algorithm that can handle streams with repeated edges, and graph structures that evolve over time. * An algorithm for the task of maintaining a weighted subsample of items in a stream, when the items must be sampled according to their weight, and the weights are dynamically changing. To our knowledge, this is the first such algorithm designed for dynamically evolving weights. We expect it may be useful as a building block for other streaming algorithms on dynamic data sets.
Source term estimation using noble gas and aerosol samples
Algorithms that estimate the location, time, and magnitude of a point-source atmospheric release using remotely sampled air concentrations typically use data for a single chemical or radioactive isotope. Here, a Bayesian algorithm is presented that uses data from multiple radioactive isotopes that are all released in the same short-duration event. Data from noble gas and aerosol samplers can be used simultaneously in the model. Application to a large synthetic data set using four isotopes shows the new algorithm generally gives more accurate location and time estimates than a comparable model using a single isotope.
Bayesian learning of orthogonal embeddings for multi-fidelity Gaussian Processes
Uncertainty propagation in complex engineering systems often poses significant computational challenges related to modeling and quantifying probability distributions of model outputs, as those emerge as the result of various sources of uncertainty that are inherent in the system under investigation. Gaussian Processes regression (GPs) is a robust meta-modeling technique that allows for fast model prediction and exploration of response surfaces. Multi-fidelity variations of GPs further leverage information from cheap and low fidelity model simulations in order to improve their predictive performance on the high fidelity model. In order to cope with the high volume of data required to train GPs in high dimensional design spaces, a common practice is to introduce latent design variables that are typically projections of the original input space to a lower dimensional subspace, and therefore substitute the problem of learning the initial high dimensional mapping, with that of training a GP on a low dimensional space. Here in this paper, we present a Bayesian approach to identify optimal transformations that map the input points to low dimensional latent variables. The \projection" mapping consists of an orthonormal matrix that is considered a priori unknown and needs to be inferred jointly with the GP parameters, conditioned on the available training data. The proposed Bayesian inference scheme relies on a two-step iterative algorithm that samples from the marginal posteriors of the GP parameters and the projection matrix respectively, both using Markov Chain Monte Carlo (MCMC) sampling. In order to take into account the orthogonality constraints imposed on the orthonormal projection matrix, a Geodesic Monte Carlo sampling algorithm is employed, that is suitable for exploiting probability measures on manifolds. We extend the proposed framework to multi-fidelity models using GPs including the scenarios of training multiple outputs together. We validate our framework on three synthetic problems with a known lower-dimensional subspace. The benefits of our proposed framework, are illustrated on the computationally challenging aerodynamic optimization of a last-stage blade for an industrial gas turbine, where we study the effect of an 85-dimensional shape parameterization of a three-dimensional airfoil on two output quantities of interest, specifically on the aerodynamic efficiency and the degree of reaction
Manifold Sampling for Optimizing Nonsmooth Nonconvex Compositions
Here we propose a manifold sampling algorithm for minimizing a nonsmooth composition $f= h\circ F$, where we assume $h$ is nonsmooth and may be inexpensively computed in closed form and $F$ is smooth but its Jacobian may not be available. We additionally assume that the composition $h\circ F$ defines a continuous selection. Manifold sampling algorithms can be classified as model-based derivative-free methods, in that models of $F$ are combined with particularly sampled information about $h$ to yield local models for use within a trust-region framework. We demonstrate that cluster points of the sequence of iterates generated by the manifold sampling algorithm are Clarke stationary. We consider the tractability of three particular subproblems generated by the manifold sampling algorithm and the extent to which inexact solutions to these subproblems may be tolerated. Numerical results demonstrate that manifold sampling as a derivative-free algorithm is competitive with state-of-the-art algorithms for nonsmooth optimization that utilize first-order information about $f$.
Robust Importance Sampling for Bayesian Model Calibration with Spatio-Temporal Data
This paper addresses two challenges in Bayesian calibration: 1) computational speed of existing sampling algorithms, and 2) calibration with spatio-temporal responses. The commonly used Markov Chain Monte Carlo (MCMC) approaches require many sequential model evaluations making the computational expense prohibitive. This paper proposes an efficient sampling algorithm: iterative importance sampling with genetic algorithm (IISGA). While iterative importance sampling enables computational efficiency, the genetic algorithm enables robustness by preventing sample degeneration and avoids getting stuck in multimodal search spaces. An inflated likelihood further enables robustness in high-dimensional parameter spaces by enlarging the target distribution. Spatio-temporal data complicate both surrogate modeling, which is necessary for expensive computational models, and the likelihood estimation. In this work, singular value decomposition is investigated for reducing the high-dimensional field data to a lower-dimensional space prior to Bayesian calibration. Then the likelihood is formulated and Bayesian inference is performed in the lower-dimension, latent space. An illustrative example is provided to demonstrate IISGA relative to existing sampling methods, and then IISGA is employed to calibrate a thermal battery model with 26 uncertain calibration parameters and spatio-temporal response data.
The combined force field-sampling problem in simulations of disordered amyloid-β peptides
Molecular dynamics simulations of intrinsically disordered proteins (IDPs) can provide high resolution structural ensembles if the force field is accurate enough and if the simulation sufficiently samples the conformational space of the IDP with the correct weighting of sub-populations. Here, we investigate the combined force field–sampling problem by testing a standard force field as well as newer fixed charge force fields, the latter specifically motivated for better description of unfolded states and IDPs, and comparing them with a standard temperature replica exchange (TREx) protocol and a non-equilibrium Temperature Cool Walking (TCW) sampling algorithm. The force field and sampling combinations are used to characterize the structural ensembles of the amyloid-beta peptides Aβ42 and Aβ43, which both should be random coils as shown recently by experimental nuclear magnetic resonance (NMR) and 2D Förster resonance energy transfer (FRET) experiments. The results illustrate the key importance of the sampling algorithm: while the standard force field using TREx is in poor agreement with the NMR J-coupling and nuclear Overhauser effect and 2D FRET data, when using the TCW method, the standard and optimized protein-water force field combinations are in very good agreement with the same experimental data since the TCW sampling method produces qualitatively different ensembles than TREx. Lastly, we also discuss the relative merit of the 2D FRET data when validating structural ensembles using the different force fields and sampling protocols investigated in this work for small IDPs such as the Aβ42 and Aβ43 peptides.
MCNP 6.2.0 Delayed-Particle Production Improvements
Development began on a delayed-particle physics package for the Monte Carlo Radiation transport code MCNPX in 2004 incorporating the CINDER90 depletion code and delayed particle libraries. When MCNP5 was merged with MCNPX to form MCNP6 in 2010, this delayed-particle capability was passed on and has been continuously improved in subsequent versions of MCNP6. Today, the delayed-particle feature in MCNP6 allows for the production of delayed neutrons, gammas, betas, and alpha particles. Future development includes the addition of delayed-positron production in the next release of MCNP6. Delayed particles can be sampled with MCNP using the ACT card in conjunction with the FISSION and NONFISS keywords. The user can specify one or more of the four available delayed-particle types mentioned above on one or both of the FISSION and NONFISS keywords. If a delayed-particle type is selected on the FISSION keyword, then delayed production of that particle from fission products and their progeny will be produced. Likewise, if a delayed-particle type is selected on the NONFISS keyword then delayed production of that particle from non-fission interactions and spontaneous decay will be produced. In some cases, the sheer number of delayed particles that can be produced from nuclear reactions and spontaneous decay (i.e. hundreds of delayed gamma lines when spontaneous fission is possible) can make it difficult to achieve converged solutions in an efficient and timely manner. Truncation methods are the simplest form of variance reduction and can be used to decrease convergence times by limiting calculations to the phase space of interest. This paper introduces several such truncation methods that allow the user to restrict the number of possible delayed particles available for sampling to only those of interest in the problem. It is important to stress that the use of any truncation method should be verified by first assessing the impact of the truncated particles (i.e. by disabling the truncation) for one or more baseline cases. In addition to these new methods, a new delayed-gamma sampling algorithm implemented in MCNP 6.2.0 is introduced. Comparisons to ENDF data and analytical benchmarking with the new algorithm show a substantial improvement in accuracy over previous versions. New improvements to the delayed-particle capability have been added to MCNP 6.2.0. These improvements include an updated treatment of the THRESH keyword which omits delayed-particle lines below a set amplitude, and addition of the PECUT and HLCUT keywords which allow the user to truncate delayed-particles lines based on energy and half-life, respectively. The delayed-gamma sampling algorithm has also been improved to allow exact energy sampling. Comparisons of MCNP calculations directly to ENDF data and analytical benchmarks have verified the significant improvement in the new delayed gamma sampling algorithm. (authors)
Efficiently sampling conformations and pathways using the concurrent adaptive sampling (CAS) algorithm
Molecular dynamics (MD) simulations are useful in obtaining thermodynamic and kinetic properties of bio-molecules but are limited by the timescale barrier, i.e., we may be unable to efficiently obtain properties because we need to run microseconds or longer simulations using femtoseconds time steps. While there are several existing methods to overcome this timescale barrier and efficiently sample thermodynamic and/or kinetic properties, problems remain in regard to being able to sample un- known systems, deal with high-dimensional space of collective variables, and focus the computational effort on slow timescales. Hence, in this work, a new sampling method, called the “Concurrent Adaptive Sampling (CAS) algorithm,” has been developed to tackle these three issues and efficiently obtain conformations and pathways. The method is not constrained to use only one or two collective variables, unlike most reaction coordinate-dependent methods. Instead, it can use a large number of collective vari- ables and uses macrostates (a partition of the collective variable space) to enhance the sampling. The exploration is done by running a large number of short simula- tions, and a clustering technique is used to accelerate the sampling. In this paper, we introduce the new methodology and show results from two-dimensional models and bio-molecules, such as penta-alanine and triazine polymer
Improved Nested Sampling and Surrogate-Enabled Comparison With Other Marginal Likelihood Estimators
Estimating marginal likelihood is of central importance to Bayesian model selection and/or model averaging. The nested sampling method has been recently used together with the Metropolis-Hasting (M-H) sampling algorithm for estimating marginal likelihood. This study develops a new implementation of nested sampling by using the DiffeRential Evolution Adaptive Metropolis (DREAMzs) sampling algorithm. The two implementations of nested sampling are evaluated for four models of a synthetic groundwater flow modeling. The DREAMzs-based nested sampling outperforms the M-H-based nested sampling in terms of their accuracy, convergence, efficiency, and stability, which is attributed to the fact that DREAMzs is more robust than M-H for parameter sampling. The nested sampling method is also compared with four other methods (arithmetic mean, harmonic mean, stabilized harmonic mean, and thermodynamic integration) commonly used for estimating marginal likelihood and posterior probability of the four groundwater models. The comparative study requires hundreds of millions of model executions, which would not be possible without using surrogate models to replace the original models. Using the arithmetic mean estimator as the reference, the comparison reveals that thermodynamic integration outperforms nested sampling in terms of accuracy and stability, whereas nested sampling is more computationally efficient to reach to convergence. Here, the harmonic mean and stabilized harmonic mean methods give the worst marginal likelihood estimation. Accurate marginal likelihood estimation is important for accurate estimation of posterior model probability and better predictive performance of Bayesian model averaging.
Algorithms in Diffraction Profile Analysis
This chapter introduces several approaches to crystallographic structure refinement from diffraction profile data including the Rietveld method, global optimization methods, and Bayesian modeling. First, the Rietveld method and related nonlinear least squares approaches to analyzing diffraction profiles are briefly discussed. We then review the most common algorithms within global optimization as alternative methods of structure solution: grid search, simulated annealing, and genetic algorithms. As the Rietveld method is the standard tool for analyzing diffraction profile data, we catalogue its challenges and limitations. A Bayesian approach to diffraction profile analysis is presented as an alternative, along with several sampling algorithms including Gibbs sampling, variants of Metropolis sampling, Hamiltonian Monte Carlo, and approximate Bayesian computing. These sampling algorithms are applied to the fitting of a neutron diffraction profile from a National Institute of Standards and Technology silicon standard reference material, and their results are compared.
Practical rare event sampling for extreme mesoscale weather
Extreme mesoscale weather, including tropical cyclones, squall lines, and floods, can be enormously damaging and yet challenging to simulate; hence, there is a pressing need for more efficient simulation strategies. Here, we present a new rare event sampling algorithm called quantile diffusion Monte Carlo (quantile DMC). Quantile DMC is a simple-to-use algorithm that can sample extreme tail behavior for a wide class of processes. We demonstrate the advantages of quantile DMC compared to other sampling methods and discuss practical aspects of implementing quantile DMC. To test the feasibility of quantile DMC for extreme mesoscale weather, we sample extremely intense realizations of two historical tropical cyclones, 2010 Hurricane Earl and 2015 Hurricane Joaquin. Our results demonstrate quantile DMC’s potential to provide low-variance extreme weather statistics while highlighting the work that is necessary for quantile DMC to attain greater efficiency in future applications.
Rejection Sampling with Autodifferentiation -- Case study: Fitting a Hadronization Model
We present an autodifferentiable rejection sampling algorithm termed Rejection Sampling with Autodifferentiation (RSA). In conjunction with reweighting, we show that RSA can be used for efficient parameter estimation and model exploration. Additionally, this approach facilitates the use of unbinned machine-learning-based observables, allowing for more precise, data-driven fits. To showcase these capabilities, we apply an RSA-based parameter fit to a simplified hadronization model.
Reactive Monte Carlo sampling with an ab initio potential
We present the first application of reactive Monte Carlo in a first-principles context. The algorithm samples in a modified NVT ensemble in which the volume, temperature, and total number of atoms of a given type are held fixed, but molecular composition is allowed to evolve through stochastic variation of chemical connectivity. We discuss general features of the method, as well as techniques needed to enhance the efficiency of Boltzmann sampling. Finally, we compare the results of simulation of NH{sub 3} to those of ab initio molecular dynamics (AIMD). We find that there are regions of state space for which RxMC sampling is much more efficient than AIMD due to the “rare-event” character of chemical reactions.
Classical Simulation of Boson Sampling Based on Graph Structure
Boson sampling is a fundamentally and practically important task that can be used to demonstrate quantum supremacy using noisy intermediate-scale quantum devices. In this Letter, we present classical sampling algorithms for single-photon and Gaussian input states that take advantage of a graph structure of a linear-optical circuit. The algorithms’ complexity grows as so-called treewidth, which is closely related to the connectivity of a given linear-optical circuit. Using the algorithms, we study approximated simulations for local Haar-random linear-optical circuits. For equally spaced initial sources, we show that, when the circuit depth is less than the quadratic in the lattice spacing, the efficient simulation is possible with an exponentially small error. Notably, right after this depth, photons start to interfere each other and the algorithms’ complexity becomes subexponential in the number of sources, implying that there is a sharp transition of its complexity. Finally, when a circuit is sufficiently deep enough for photons to typically propagate to all modes, the complexity becomes exponential as generic sampling algorithms. We numerically implement a likelihood test with a recent Gaussian boson sampling experiment and show that the treewidth-based algorithm with a limited treewidth renders a larger likelihood than the experimental data.