Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “algorithm selection”

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.

At least 433 records · Page 24

The Gemini NICI planet-finding campaign: The companion detection pipeline

We present high-contrast image processing techniques used by the Gemini NICI Planet-Finding Campaign to detect faint companions to bright stars. The Near-Infrared Coronographic Imager (NICI) is an adaptive optics instrument installed on the 8 m Gemini South telescope, capable of angular and spectral difference imaging and specifically designed to image exoplanets. The Campaign data pipeline achieves median contrasts of 12.6 mag at 0.''5 and 14.4 mag at 1'' separation, for a sample of 45 stars (V = 4.3-13.9 mag) from the early phase of the campaign. We also present a novel approach to calculating contrast curves for companion detection based on 95% completeness in the recovery of artificial companions injected into the raw data, while accounting for the false-positive rate. We use this technique to select the image processing algorithms that are more successful at recovering faint simulated point sources. We compare our pipeline to the performance of the Locally Optimized Combination of Images (LOCI) algorithm for NICI data and do not find significant improvement with LOCI.

79 ASTRONOMY AND ASTROPHYSICS↗

Dark Energy Survey Year 3 results: galaxy clustering and systematics treatment for lens galaxy samples

ABSTRACT In this work, we present the galaxy clustering measurements of the two DES lens galaxy samples: a magnitude-limited sample optimized for the measurement of cosmological parameters, maglim, and a sample of luminous red galaxies selected with the redmagic algorithm. maglim/redmagic sample contains over 10 million/2.5 million galaxies and is divided into six/five photometric redshift bins spanning the range z ∈ [0.20, 1.05]/z ∈ [0.15, 0.90]. Both samples cover 4143 $\deg ^2$ over which we perform our analysis blind, measuring the angular correlation function with an S/N ∼ 63 for both samples. In a companion paper, these measurements of galaxy clustering are combined with the correlation functions of cosmic shear and galaxy–galaxy lensing of each sample to place cosmological constraints with a 3 × 2pt analysis. We conduct a thorough study of the mitigation of systematic effects caused by the spatially varying survey properties and we correct the measurements to remove artificial clustering signals. We employ several decontamination methods with different configurations to ensure the robustness of our corrections and to determine the systematic uncertainty that needs to be considered for the final cosmology analyses. We validate our fiducial methodology using lognormal mocks, showing that our decontamination procedure induces biases no greater than 0.5σ in the (Ωm, b) plane, where b is the galaxy bias.

79 ASTRONOMY AND ASTROPHYSICS↗

Dark Energy Survey year 1 results: Galaxy clustering for combined probes

We measure the clustering of DES year 1 galaxies that are intended to be combined with weak lensing samples in order to produce precise cosmological constraints from the joint analysis of large-scale structure and lensing correlations. Two-point correlation functions are measured for a sample of 6.6×10 5 luminous red galaxies selected using the redMaGiC algorithm over an area of 1321 square degrees, in the redshift range 0.15 z /(1+z)=0.017. We quantify and correct spurious correlations induced by spatially variable survey properties, testing their impact on the clustering measurements and covariance. We demonstrate the sample’s robustness by testing for stellar contamination, for potential biases that could arise from the systematic correction, and for the consistency between the two-point auto- and cross-correlation functions. We show that the corrections we apply have a significant impact on the resultant measurement of cosmological parameters, but that the results are robust against arbitrary choices in the correction method. We find the linear galaxy bias in each redshift bin in a fiducial cosmology to be b(σ 8 /0.81)| z=0.24 =1.40±0.07, b(σ 8 /0.81)| z=0.38 =1.60±0.05, b(σ 8 /0.81)| z=0.53 =1.60±0.04 for galaxies with luminosities L/L * >0.5, b(σ 8 /0.81)| z=0.68 =1.93±0.04 for L/L * >1 and b(σ 8 /0.81)| z=0.83 =1.98±0.07 for L/L * >1.5, broadly consistent with expectations for the redshift and luminosity dependence of the bias of red galaxies. We show these measurements to be consistent with the linear bias obtained from tangential shear measurements.

79 ASTRONOMY AND ASTROPHYSICS↗

Detecting Important Drivers of Gridded Population Modeling With Machine Learning

High-resolution population datasets have been lever-aged across a broad swath of domains, such as climate change, public policy, humanitarian aid, and rescue operations, among others. Machine learning methods were adopted to generate high-resolution or gridded population estimates by using various geospatial input features such as buildings, roads, and nighttime lights. In this study, we evaluate the importance of population features using Random Forest models across three levels of analysis, utilizing permutation measures. Our research aims to address key questions to enhance our understanding of high-resolution population modeling, such as: Are certain features globally (10 countries collectively) more important than others? Do optimal features vary by country? Within each country, do feature importance differ across administrative units? What similarities exist in feature importance at the global, country, and administrative unit levels? To answer these questions, we leverage the Kneedle algorithm to automate the selection of optimum features. We find that there are patterns displayed by features across spatial boundaries, evidenced by the same feature being the most important indicator of population across 7 of the 10 countries modeled. Our findings indicate that while important features may vary across geographies, certain features consistently hold greater importance than others agnostic of geography.

Lebakula, Viswadeep [ORNL] (ORCID:0000000152935914↗

Temporal ensemble learning of univariate methods for short term load forecasting

Short term load forecasting (STLF) is a fundamental component of demand response programs in smart grids. Recent literature has focused on complex neural networks and deep learning models for solving STLF. While these models work well for load forecasting with complex non-linear relationships, they have been shown to be less effective than simpler univariate models for STLF problems with under a 6 hours horizon. Given the lack of multivariate data (such as temperature) in many practical datasets, we need better univariate prediction models for STLF. By partitioning the dataset by temporal features, we develop a novel ensemble learning method that exploits daily seasonality in electricity consumption data to improve accuracy of popular univariate models. We train an ensemble of models from the dataset partitions. We develop a variety of methods, including Ridge Regression, to increase the robustness of the ensemble prediction. To show the effectiveness of our approach, we perform detailed evaluation using an aggregated user electricity consumption dataset collected by the Los Angeles Department of Water and Power (LADWP). We select four popular prediction algorithms in literature for our experiments, including Kernel Regression (KR), Support Vector Regression (SVR) and neural network approaches. We compare the performance of these algorithms applying our ensemble approach to training only one single model. Our approach leads to an 11.2% decrease in mean absolute percentage error (MAPE) and 21.3% decrease in root mean squared error (RMSE) over the single model approach for KR, and a 30% and 32.4% decrease in MAPE and RMSE respectively for SVR. These ensemble models also outperform the neural network approaches.

14 SOLAR ENERGY↗

Sequential Perturbation-based FDIA Detection using DERs for Unbalanced Distribution System

The power distribution system with its reliance on automated monitoring and control infrastructure makes it a complicated cyber-physical system and vulnerable to various cyber-attacks. This paper proposes a moving target defense inspired false data injection attack (FDIA) detection mechanism for an unbalanced distribution system. In the proposed grid diagnosis service framework, the distribution system operator judiciously manipulates power outputs of a subset of inverter-based distribution energy resources (DERs) to create secret low magnitude voltage perturbations which are inconsequential to the normal operation of the grid. A mixed-integer-linear-programming algorithm is developed to select the optimal set of DERs that can create a voltage perturbation signal of the required magnitude at each sensor location at a minimum cost. Then, a sequential detector is applied that detects for the FDIA as measurements are received from individual sensors. The performance of the proposed perturbation-based FDIA detection framework is demonstrated via simulation of the IEEE 123 bus test system.

Attack Detection↗

An Experiment-based Distribution Level Performance Comparison among PMUs

This paper presents a experiment-based distribution level performance comparison among three Phasor Measurement Units (PMUs). Several evaluation criteria, including the total vector error, the phase angle error, the frequency error, the rate of change of frequency, the response time, the settling time, the overshoot, and the algorithm window size, are selected to compare the static and dynamic performances of the PMUs under steady state and step response test conditions. In order to have a more realistic test environment, a test is setup in which PMUs under evaluation have exactly the same input signals. The quantitative experiment result analysis gives an end-user guideline to the PMU selection regarding distribution level applications.

Yin, He↗

eFESTA: Ensemble Feature Exploration with Surface Density Estimates

We propose surface density estimate (SDE) to model the spatial distribution of surface features-isosurfaces, ridge surfaces, and streamsurfaces-in 3D ensemble simulation data. The inputs of SDE computation are surface features represented as polygon meshes, and no field datasets are required (e.g., scalar fields or vector fields). The SDE is defined as the kernel density estimate of the infinite set of points on the input surfaces and is approximated by accumulating the surface densities of triangular patches. We also propose an algorithm to guide the selection of a proper kernel bandwidth for SDE computation. An ensemble Feature Exploration method based on Surface densiTy EstimAtes (eFESTA) is then proposed to extract and visualize the major trends of ensemble surface features. For an ensemble of surface features, each surface is first transformed into a density field based on its contribution to the SDE, and the resulting density fields are organized into a hierarchical representation based on the pairwise distances between them. The hierarchical representation is then used to guide visual exploration of the density fields as well as the underlying surface features. We demonstrate the application of our method using isosurface in ensemble scalar fields, Lagrangian coherent structures in uncertain unsteady flows, and streamsurfaces in ensemble fluid flows.

Computational modeling↗

Global Pathway Selection with Zero-RK

Global Pathway Selection (GPS) is an algorithm to effectively generates reduced (skeletal) chemistry mechanisms, which speeds up simulations and can be used as a systematic analytics tool to extract insights from complex reacting system. This release is an extension of the original code to run in parallel and to use LLNL's Zero-RK solver for fast solution of chemical problems.

Whitesides, RussellA↗

Global Pathway Selection with Zero-RK v0.5

Global Pathway Selection (GPS) is an algorithm to effectively generates reduced (skeletal) chemistry mechanisms, which speeds up simulations and can be used as a systematic analytics tool to extract insights from complex reacting system. This release is an extension of the original code to run in parallel and to use LLNL's Zero-RK solver for fast solution of chemical problems.

Whitesides, RussellA [Lawrence Livermore National ↗

Estimating Sparse Direct Effects in Multivariate Regression With the Spike-and-Slab LASSO

The multivariate regression interpretation of the Gaussian chain graph model simultaneously parametrizes (i) the direct effects of p predictors on q outcomes and (ii) the residual partial covariances between pairs of outcomes. We introduce a new method for fitting sparse versions of these models with spike-and-slab LASSO (SSL) priors. We develop an Expectation Conditional Maximization algorithm to obtain sparse estimates of the p × q matrix of direct effects and the q × q residual precision matrix. Our algorithm iteratively solves a sequence of penalized maximum likelihood problems with self-adaptive penalties that gradually filter out negligible regression coefficients and partial covariances. Because it adaptively penalizes individual model parameters, our method is seen to outperform fixed-penalty competitors on simulated data. We establish the posterior contraction rate for our model, buttressing our method’s excellent empirical performance with strong theoretical guarantees. Using our method, we estimated the direct effects of diet and residence type on the composition of the gut microbiome of elderly adults.

EM algorithm↗

Designing and Testing Broadly-Protective Filoviral Vaccines Optimized for Cytotoxic T-Lymphocyte Epitope Coverage

We report the rational design and in vivo testing of mosaic proteins for a polyvalent pan-filoviral vaccine using a computational strategy designed for the Human Immunodeficiency Virus type 1 (HIV-1) but also appropriate for Hepatitis C virus (HCV) and potentially other diverse viruses. Mosaics are sets of artificial recombinant proteins that are based on natural proteins. The recombinants are computationally selected using a genetic algorithm to optimize the coverage of potential cytotoxic T lymphocyte (CTL) epitopes. Because evolutionary history differs markedly between HIV-1 and filoviruses, we devised an adapted computational technique that is effective for sparsely sampled taxa; our first significant result is that the mosaic technique is effective in creating high-quality mosaic filovirus proteins. The resulting coverage of potential epitopes across filovirus species is superior to coverage by any natural variants, including current vaccine strains with demonstrated cross-reactivity. The mosaic cocktails are also robust: mosaics substantially outperformed natural strains when computationally tested against poorly sampled species and more variable genes. Furthermore, in a computational comparison of cross-reactive potential a design constructed prior to the Bundibugyo outbreak performed nearly as well against all species as an updated design that included Bundibugyo. These points suggest that the mosaic designs would be more resilient than natural-variant vaccines against future Ebola outbreaks dominated by novel viral variants. We demonstrate in vivo immunogenicity and protection against a heterologous challenge in a mouse model. This design work delineates the likely requirements and limitations on broadly-protective filoviral CTL vaccines.

59 BASIC BIOLOGICAL SCIENCES↗

Neutral Current Elastic Scattering and the Strange Spin Structure of the Proton

Neutrinos can be used as a unique and informative probe of the structure within nucleons. This thesis presents the tools and methodology for studying the strange quark spin in the nucleon through neutral current elastic neutrino-proton scattering in the MicroBooNE experiment located at the Fermi National Accelerator Lab. An automated boosted decision tree based proton identification algorithm is used to select proton in liquid argon TPC data with a 70\% efficiency. After a set of protons are selected, a logistic regression model is used to select neutral current elastic proton interactions in MicroBooNE. Neutral current elastic proton interactions are selected with an 11\% efficiency and 30\% purity. The number of selected events measured in MicroBooNE data as a function of $Q^2$ is compared directly to the number of selected events in MicroBooNE simulation. An event reweighting scheme is used to vary the expected number of eventsin MicroBooNE simulation based on the values o f the physics parameters of interest. The likelihood for any given value of the physical parameters can then be easily calculated and the probability distributions of the physics parameters can be sampled using Markov Chain Monte Carlo. Future improvements to the detector physics models in MicroBooNE are required for a precise determination of the strange quark spin structure in the proton. With the current level of uncertainty on the detector physics models, we are only able to constrain the net strange quark spin, $\Delta s$ to the range $-1.8 < \Delta s < 3.8$ with 95\% confidence.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Group Structure Machine Learning Proposal

Nuclear data is the linchpin underwriting several fundamental capabilities and mission needs at LANL. New techniques such as machine learning can be brought to bear to solve old problems such as multigroup cross-section accuracy. In neutron transport, generating multigroup cross sections is a complex and arcane task, but a crucial one, as accurate solutions require appropriate cross sections. There are two key challenges when generating multigroup cross sections: (1) choosing an accurate weight function, and (2) choosing appropriate energy boundaries. Often, energy boundaries are chosen using “expert judgment” that is not documented and is difficult to replicate. The long-standing Los Alamos 30-group structure has been in use since at least 1969 and is still in use today. Simplistic attempts over the years since to improve on the 30-group structure have been met with limited success. Machine learning algorithms would enable the selection of appropriate, problem-dependent group boundaries without an inordinate investment of scientist time. We will develop workflows and tools to enable these improved group boundary choices, which will reduce uncertainty and increase predictive capability of neutron-transport applications at LANL.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Progress Toward the First Search for Bound Neutron Oscillation into Antineutron in a Liquid Argon TPC

This note presents current progress for a neutron-antineutron oscillation ($n–\overline{n}$) search in MicroBooNE paving the way for the first search analysis of such process in a Liquid Argon Time Projection Chamber (LArTPC). Convolutional Neural Network (CNN) and Boosted Decision Tree (BDT) algorithms were used to select signal $n–\overline{n}$ events over cosmogenic backgrounds. The CNN-only, BDT-only, and the combined (CNN+BDT) methods were demonstrated on the Monte-Carlo signal and background events. Validation of the CNNonly and the BDT-only methods was carried out on a small dataset of MicroBooNE Run1 off-beam data, setting the starting point toward further improvement of the analysis.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Teko Usage in Aria

Demonstration of Teko preconditioning capability in Aria. Proposed future development work to aid preconditioner selection via simple heuristic algorithms is presented. Some highlight applications of Teko are included. These contain brief physics descriptions, solver performance, and solver convergence information details. Notably, no geometric information or otherwise sensitive information is provided.

Phillips, Malachi↗

Control of the Deep Grinding of Gas-Turbine Components

The efficiency of deep grinding may be increased by regulating the longitudinal supply of the table. The selection of the control algorithm is considered in machining the components of gas-turbine engines.

42 ENGINEERING↗

Hierarchical Modeling to Enhance Spectrophotometry Measurements—Overcoming Dynamic Range Limitations for Remote Monitoring of Neptunium

A robust hierarchical model has been demonstrated for monitoring a wide range of neptunium concentrations (0.75–890 mM) and varying temperatures (10–80 °C) using chemometrics and feature selection. The visible–near infrared electronic absorption spectrum (400–1700 nm) of monocharged neptunyl dioxocation (Np(V) = NpO2+) includes many bands, which have molar absorption coefficients that differ by nearly 2 orders of magnitude. The shape, position, and intensity of these bands differ with chemical interactions and changing temperature. These challenges make traditional quantification by univariate methods unfeasible. Measuring Np(V) concentration over several orders of magnitude would typically necessitate cells with varying path length, optical switches, and/or multiple spectrophotometers. Alternatively, the differences in the molar extinction coefficients for multiple absorption bands can be used to quantify Np(V) concentration over 3 orders of magnitude with a single optical path length (1 mm) and a hierarchical multivariate model. In this work, principal component analysis was used to distinguish the concentration regime of the sample, directing it to the relevant partial least squares regression submodels. Each submodel was optimized with unique feature selection filters that were selected by a genetic algorithm to enhance predictions. Through this approach, the percent root mean square error of prediction values were ≤1.05% for Np(V) concentrations and ≤4% for temperatures. This approach may be applied to other nuclear fuel cycle and environmental applications requiring real-time spectroscopic measurements over a wide range of conditions.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗