Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Inference”

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 685 records · Page 38

Rapid Inference of Logic Gate Neural Networks for Anomaly Detection in High Energy Physics

The increasing data rates and complexity of detectors at the Large Hadron Collider (LHC) necessitate fast and efficient machine learning models, particularly for rapid selection of what data to store, known as triggering. Building on recent work in differentiable logic gates, we present a public implementation of a Convolutional Differentiable Logic Gate Neural Network (CLGN). We apply this to detecting anomalies at the Level-1 Trigger at CMS using public data from the CICADA project. We demonstrate that the CLGN achieves physics performance on par with or superior to conventional quantized neural networks. We also synthesize an LGN for a Field-Programmable Gate Array (FPGA) and show highly promising FPGA characteristics, notably zero Digital Signal Processor (DSP) resource usage. This work highlights the potential of logic gate networks for high-speed, on-detector inference in High Energy Physics and beyond.

FOS: Physical sciences↗

Errors in nanoparticle growth rates inferred from measurements in chemically reacting aerosol systems

In systems in which aerosols are being formed by chemical transformations,individual particles grow due to the addition of molecular species. Efforts to improve our understanding of particle growth often focus on attempts to reconcile observed growth rates with values calculated from models. However,because it is typically not possible to measure the growth rates of individual particles in chemically reacting systems, they must be inferred from measurements of aerosol properties such as size distributions, particle number concentrations, etc. Furthermore this work discusses errors in growth rates obtained using methods that are commonly employed for analyzing atmospheric data. We analyze “data” obtained by simulating the formation of aerosols in a system in which a single chemical species is formed at a constant rate, R . We show that the maximum overestimation error in measured growth rates occurs for collision-controlled nucleation in a single-component system in the absence of a preexisting aerosol, wall losses, evaporation or dilution,as this leads to the highest concentrations of nucleated particles. Those high concentrations lead to high coagulation rates that cause the nucleation mode to grow faster than would be caused by vapor condensation alone. We also show that preexisting particles, when coupled with evaporation, can significantly decrease the concentration of nucleated particles. This can lead to decreased discrepancies between measured growth rate and true growth rate by reducing coagulation among nucleated particles. However, as particle sink processes become stronger, measured growth rates can potentially be lower than true particle growth rates. We briefly discuss nucleation scenarios in which the observed growth rate approaches zero while the true growth rate does not.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Inferred inflow forecast horizons guiding reservoir release decisions across the United States

Abstract. Medium- to long-range forecasts often guide reservoir release decisions to support water management objectives, including mitigating flood and drought risks. While there is a burgeoning field of science targeted at improving forecast products and associated decision support models, data describing how and when forecasts are applied in practice remain undeveloped. This lack of knowledge may prevent hydrological modelers from developing accurate reservoir release schemes for large-scale, distributed hydrology models that are increasingly used to assess the vulnerabilities of large regions to hydrological stress. We address this issue by estimating seasonally varying, regulated inflow forecast horizons used in the operations of more than 300 dams throughout the conterminous United States (CONUS). For each dam, we take actual forward observed inflows (perfect foresight) as a proxy for forecasted flows available to the operator and then identify for each week of the year the forward horizon that best explains the release decisions taken. Resulting “horizon curves” specify for each dam the inferred inflow forecast horizon as a function of the week of the water year. These curves are analyzed for strength of evidence for contribution of medium- to long-range forecasts in decision making. We use random forest classification to estimate that approximately 80 % of large dams and reservoirs in the US (1553±50 out of 1927 dams with at least 10 Mm3 storage capacity) adopt medium- to long-range inflow forecasts to inform release decisions during at least part of the water year. Long-range forecast horizons (more than 6 weeks ahead) are detected in the operations of reservoirs located in high-elevation regions of the western US, where snowpack information likely guides the release. A simulation exercise conducted on four key western US reservoirs indicates that forecast-informed models of reservoir operations may outperform models that neglect the horizon curve – including during flood and drought conditions.

54 ENVIRONMENTAL SCIENCES↗

Stream water sourcing from high-elevation snowpack inferred from stable isotopes of water: a novel application of d-excess values

About 80 % of the precipitation at the Colorado River's headwaters is snow, and the resulting snowmelt-driven hydrograph is a crucial water source for about 40 million people. Snowmelt from alpine and subalpine snowpack contributes substantially to groundwater recharge and river flow. However, the dynamics of snowmelt progression are not well understood because observations of the high-elevation snowpack are difficult due to challenging access in complex mountainous terrain as well as the cost and labor intensity of currently available methods. We present a novel approach to infer the processes and dynamics of high-elevation snowmelt contributions predicated upon stable hydrogen and oxygen isotope ratios observed in streamflow. We show that deuterium-excess (d-excess) values of stream water could serve as a comparatively cost-effective proxy for a catchment-integrated signal of high-elevation snowmelt contributions to catchment runoff. We sampled stable hydrogen and oxygen isotope ratios of the precipitation, snowpack, and stream water in the East River, a headwater catchment of the Colorado River, and the stream water of larger catchments at sites on the Gunnison River and Colorado River. The d-excess of snowpack increased with elevation; the upper subalpine and alpine snowpack (> 3200 m) had substantially higher d-excess compared to lower elevations (< 3200 m) in the study area. The d-excess values of stream water reflected this because d-excess values increased as the higher-elevation snowpack contributed more to stream water generation later in the snowmelt/runoff season. End-member mixing analyses based on the d-excess data showed that the share of high-elevation snowmelt contributions within the snowmelt hydrograph was on average 44 % and generally increased during melt period progression, up to 70 %. The observed pattern was consistent during 6 years for the East River, and a similar relation was found for the larger catchments on the Gunnison and Colorado rivers. High-elevation snowpack contributions were found to be higher for years with lower snowpack and warmer spring temperatures. Thus, we conclude that the d-excess of stream water is a viable proxy to observe changes in high-elevation snowmelt contributions in catchments at various scales. Inter-catchment comparisons and temporal trends of the d-excess of stream water could therefore serve as a catchment-integrated measure to monitor if mountain systems rely on high-elevation water inputs more during snow drought compared to years of average snowpack depths.

54 ENVIRONMENTAL SCIENCES↗

Antarctic permafrost processes and antiphase dynamics of cold-based glaciers in the McMurdo Dry Valleys inferred from 10 Be and 26 Al cosmogenic nuclides

Abstract. Soil and sediment mixing and associated permafrost processes are not widely studied or understood in the McMurdo Dry Valleys of Antarctica. In this study, we investigate the stability and depositional history of near-surface permafrost sediments to ∼ 3 m depth in the Pearse and lower Wright valleys using measured cosmogenic 10Be and 26Al depth profiles. In Pearse Valley, we estimate a minimum depositional age of ∼ 74 ka for the active layer and paleoactive-layer sediments (< 0.65 m). Combined depth profile modelling of 10Be and 26Al gives a depositional age for near-surface (< 1.65 m) permafrost in Pearse Valley of 180 +20/-40 ka, implying that the deposition of permafrost sediments predates MIS 5 advances of Taylor Glacier. Deeper permafrost sediments (> 2.09 m) in Pearse Valley are thus inferred to have a depositional age of > 180 ka. At a coastal, lower-elevation site in neighbouring lower Wright Valley, 10Be and 26Al depth profiles from a second permafrost core exhibit near-constant concentrations with depth and indicate the sediments are either vertically mixed after deposition or sufficiently young so that post-depositional nuclide production is negligible relative to inheritance. 26Al/10Be concentration ratios for both depth profiles range between 4.0 and 5.2 and are all lower than the nominal surface production rate ratio of 6.75, indicating that prior to deposition, these sediments experienced complex, yet similar, exposure–burial histories. Assuming a single-cycle exposure–burial scenario, the observed 26Al/10Be ratios are equivalent to a total minimum exposure–burial history of ∼ 1.2 Myr. In proximity to the depth profile core site, we measured cosmogenic 10Be and 26Al in three granite cobbles from thin, patchy drift (Taylor 2 Drift) in Pearse Valley to constrain the timing of retreat of Taylor Glacier. Assuming simple continuous exposure, our minimum, zero-erosion exposure ages suggest Taylor Glacier partially retreated from Pearse Valley no later than 65–74 ka. The timing of retreat after 65 ka and until the Last Glacial Maximum (LGM) when Taylor Glacier was at a minimum position remains unresolved. The surface cobble ages and permafrost processes reveal Taylor Glacier advances during MIS 5 were non-erosive or mildly erosive, preserving the underlying permafrost sediments and peppering boulders and cobbles upon an older, relict surface. Our results are consistent with U/Th ages from central Taylor Valley and suggest changes in moisture delivery over Taylor Dome during MIS 5e, 5c, and 5a appear to be associated with the extent of the Ross Ice Shelf and sea ice in the Ross Sea. These data provide further evidence of antiphase behaviour through retreat of a peripheral lobe of Taylor Glacier in Pearse Valley, a region that was glaciated during MIS 5. We suggest a causal relationship of cold-based glacier advance and retreat that is controlled by an increase in moisture availability during retreat of sea ice and perhaps the Ross Ice Shelf, as well as, conversely, a decrease during times of sea ice and Ross Ice Shelf expansion in the Ross Sea.

58 GEOSCIENCES↗

Large-Scale Inference of Multivariate Regression for Heavy-Tailed and Asymmetric Data

Large-scale multivariate regression is a fundamental statistical tool with a wide range of applications. Here, this study considers the problem of simultaneously testing a large number of general linear hypotheses, encompassing covariate-effect analysis, analysis of variance, and model comparisons. The challenge that accompanies a large number of tests is the ubiquitous presence of heavy-tailed and/or highly skewed measurement noise, which is the main reason for the failure of conventional least squares-based methods. For large-scale multivariate regression, we develop a set of robust inference methods to explore data features such as heavy tailedness and skewness, which are not visible to least squares methods. The new testing procedure is based on the data-adaptive Huber regression and a new covariance estimator of regression estimates. Under mild conditions, we show that our methods produce consistent estimates of the false discovery proportion. Extensive numerical experiments and an empirical study on quantitative linguistics demonstrate the advantage of the proposed method over many state-of-the-art methods when the data are generated from heavy-tailed and/or skewed distributions.

97 MATHEMATICS AND COMPUTING↗

MANTRA: An Integral Reactor Physics Experiment to Infer Actinide Capture Cross-sections from Thorium to Californium with Accelerator Mass Spectrometry

The principle of the proposed experiment is to irradiate very pure actinide samples in the Advanced Test Reactor at INL and, after a given time, determine the amount of the different transmutation products. The determination of the nuclide densities before and after neutron irradiation will allow inference of effective neutron capture cross-sections. This approach has been used in the past and the novelty of this experiment is that the atom densities of the different transmutation products will be determined using the Accelerator Mass Spectrometry technique at the ATLAS facility located at ANL. It is currently planned to irradiate the following isotopes: 232Th, 235U, 236U, 238U, 237Np, 238Pu, 239Pu, 240Pu, 241Pu, 242Pu, 241Am, 243Am, 244Cm and 248Cm.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Statistical Inference for Big Data Problems in Molecular Biophysics

We highlight the role of statistical inference techniques in providing biological insights from analyzing long time-scale molecular simulation data. Technologi- cal and algorithmic improvements in computation have brought molecular simu- lations to the forefront of techniques applied to investigating the basis of living systems. While these longer simulations, increasingly complex reaching petabyte scales presently, promise a detailed view into microscopic behavior, teasing out the important information has now become a true challenge on its own. Mining this data for important patterns is critical to automating therapeutic intervention discovery, improving protein design, and fundamentally understanding the mech- anistic basis of cellular homeostasis.

Ramanathan, Arvind↗

Misleading or Falsification? Inferring Deceptive Strategies and Types in Online News and Social Media

Deceptive information in online news and social media has had dramatic effect on our society in recent years. This study is the first to gain deeper insights into writers' intent behind digital misinformation by analyzing psycholinguistic signals: moral foundations and connotations extracted from different types of deceptive news ranging from strategic disinformation to propaganda and hoaxes. To ensure consistency of our findings and generalizability across domains, we experiment with data from: (1) confirmed cases of disinformation in news summaries, (2) {propaganda}, hoax, and disinformation news pages, and (3) social media news. We first contrast lexical markers of biased language, syntactic and stylistic signals, and connotations across deceptive news types including disinformation, propaganda, and hoaxes, and {deceptive} strategies including misleading or falsification. We then incorporate these insights to build machine learning and deep learning predictive models to infer deception strategies and deceptive news types. Our experimental results demonstrate that unlike earlier work on deception detection, content combined with biased language markers, moral foundations, and connotations leads to better predictive performance of deception strategies compared to syntactic and stylistic signals (as reported in earlier work on deceptive reviews). Falsification strategy is easier to identify than misleading strategy. Disinformation is more difficult to predict than to propaganda or hoaxes. Deceptive news types (disinformation, propaganda, and hoaxes), unlike deceptive strategies (falsification and misleading), are more salient, and thus easier to identify in tweets than in news reports. Finally, our novel connotation analysis across deception types provides deeper understanding of writers' perspectives and therefore reveals the intentions behind digital misinformation.

deception↗

Method and system for information retrieval and aggregation from inferred user reasoning

The visual analytic system enables information retrieval within large text collections. Typically, users have to directly and explicitly query information to retrieve it. With this system and process, the reasoning of the user is inferred from the user interaction they perform in a visual analytic tool, and the appropriate information to query, process, and visualize is systematically determined.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗

Inferring Structural Parameters of Low-Surface-Brightness-Galaxies with Uncertainty Quantification using Bayesian Neural Networks

Measuring the structural parameters (size, total brightness, light concentration, etc.) of galaxies is a significant first step towards a quantitative description of different galaxy populations. In this work, we demonstrate that a Bayesian Neural Network (BNN) can be used for the inference, with uncertainty quantification, of such morphological parameters from simulated low-surface-brightness galaxy images. Compared to traditional profile-fitting methods, we show that the uncertainties obtained using BNNs are comparable in magnitude, well-calibrated, and the point estimates of the parameters are closer to the true values. Our method is also significantly faster, which is very important with the advent of the era of large galaxy surveys and big data in astrophysics.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

New Machine Learning Techniques for Simulation-Based Inference: InferoStatic Nets, Kernel Score Estimation, and Kernel Likelihood Ratio Estimation

We propose an intuitive, machine-learning approach to multiparameter inference, dubbed the InferoStatic Networks (ISN) method, to model the score and likelihood ratio estimators in cases when the probability density can be sampled but not computed directly. The ISN uses a backend neural network that models a scalar function called the inferostatic potential $\varphi$. In addition, we introduce new strategies, respectively called Kernel Score Estimation (KSE) and Kernel Likelihood Ratio Estimation (KLRE), to learn the score and the likelihood ratio functions from simulated data. We illustrate the new techniques with some toy examples and compare to existing approaches in the literature. We mention en passant some new loss functions that optimally incorporate latent information from simulations into the training procedure.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Occupant thermal comfort inference using body shape information

Occupant thermal comfort may be inferred and improved using body shape information. Height, weight, and shoulder circumference of an occupant of a room may be obtained using a depth sensor. A model may be utilized that is trained on a dataset including information reflecting of occupant comfort within the room versus temperature, the model receiving, as inputs, the height, the weight, and the shoulder circumference of the occupant and environmental information and outputting a comfort class. A temperature set-point for is identified which the room occupant is identified by the model as having the comfort class being indicative of user comfort. Heating, ventilation, and air conditioning (HVAC) controls are adjusted for the room to the identified temperature set-point.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Characteristics of SiC neutron sensor spectrum unfolding process based on Bayesian inference

This paper deals with SiC detector signal interpretation in neutron radiation measurements in mixed neutron gamma radiation fields, which is called the detector inverse problem or the spectrum unfolding, and it aims in finding a representation of the primary radiation, based on the measured detector signals. In our novel methodology we resort to Bayesian inference approach. In the developed procedure the resultant spectra is unfolded form detector channels reading, where the estimated neutron fluence in a group structure is obtained with its statistical characteristic comprising of standard deviation and correlation matrix. In the paper we present results of unfolding process for case of D-T neutron source in neutron moderating environment. Discussions of statistical properties of obtained results are presented as well as of the physical meaning of obtained correlation matrix of estimated group fluence. The presented works has been carried out within the I-SMART project, which is part of the KIC InnoEnergy R and D program. (authors)

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Bayesian inference for high confidence signal validation and sensor calibration assessment - 328

Online Monitoring (OLM) systems has been investigated for applications in the nuclear industry for the past two decades. These systems offer economic and technological advantages over the currently implemented time based testing methods for sensors and associated instrumentations. The advantages offered not only augment economic viability of current nuclear reactors but will be an inevitable change called for to suit the technologies of the advanced reactors (Gen III+). Irrespective, the acceptance of OLM into the regulatory structure in the U.S. is still hindered by technical and structural clarity on uncertainty quantification used by these systems. A Bayesian inference technique is implemented in the form of a residual model to quantify uncertainty in OLM predictions. The developed system used Gaussian processes to represent the OLM predictions and model inadequacy, the stationary representation of model based uncertainty. In addition, model selection techniques are used in conjunction with numerical methods such as Genetic Algorithm (GA) and Gibb's sampler to add robustness to the estimation model inadequacy, particularly in the presence of drifts and faulted conditions. The prediction model developed herein find applicability in fault detection, obtaining real values of processes after the onset of faults etc. Further development looks into the extension of current method for a multiple output scale implementation. (authors)

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Leveraging Hydropower Multi-Sensor Data for Inference and Age-Informed Modeling

Increased demand of operational flexibility such as faster ramp up/down in generation, and more frequent start/stops are putting hydropower plants and their associated components in unprecedented stress. Consequently, these plants are at the high risk of extended and more frequent outage to accommodate unscheduled, and unexpected maintenance. Therefore, hydropower plants are in critical need of data driven and age-informed analysis for their regular and unscheduled operation. Yet not all hydropower plants are exhaustively equipped with sensors and/or measurement streams for their respective components – demanding solutions on how to detect, identify, and locate the cause of any event from the unobservable. Idaho National Laboratory (INL) analyzed the anonymized measurements and event records from the Hydropower Research Institute (HRI) to address this issue, as part of the Water Power Technologies Office (WPTO) funded one year multi-lab project. First, we investigated how time series of multiple sensor measurements can be leveraged to identify an event “root cause” as well as to develop an inference (i.e., estimate the unobservable) problem. INL also investigated how individual hydropower components’ reaction or response times vary across the pre-event, during event, and post-event conditions – enabling the hydropower dynamic models to be age-informed. Finally, the impact of clustering multi-sensor time series on short-term vibration prediction is analyzed. INL will present key findings from these analyses and recommend next steps for stakeholder adoption.

13 HYDRO ENERGY↗

Inferring UAV Position Relative to Smartphone Sensor Based on Acoustic Characteristics

As commercial and non-commercial unmanned aerial vehicles (UAVs) improve technologically and increase in accessibility, their potential for use in actions antithetical to nuclear non-proliferation efforts increases as well. The ability to identify and locate UAVs using ubiquitous, inexpensive hardware and software would be advantageous as both a deterrent and an initial warning system. Our team looks to utilize smartphones as acoustic sensors to record and characterize UAV movement based on the sound produced while in flight. Multi-rotor UAVs produce identifiable acoustic spectral characteristics with frequencies and harmonics corresponding to motor revolutions per minute and the propeller characteristics. As the source moves closer to the receiver these spectral patterns remain relatively stable (at slow speeds), while the energy observed from the waves increases. The goal of this work is to utilize a small network of smartphone sensors with known, fixed positions to infer the relative position of an in-flight UAV with respect to the position of the smartphones through the variation in acoustic energy.

46 - INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AN↗

Automated Membership Inference Attacks: Discovering MIA Signal Computations using LLM Agents

Membership inference attacks (MIAs), which enable adversaries to determine whether specific data points were part of a model's training dataset, have emerged as an important framework to understand, assess, and quantify the potential information leakage associated with machine learning systems. Designing effective MIAs is a challenging task that usually requires extensive manual exploration of model behaviors to identify potential vulnerabilities. In this paper, we introduce AutoMIA -- a novel framework that leverages large language model (LLM) agents to automate the design and implementation of new MIA signal computations. By utilizing LLM agents, we can systematically explore a vast space of potential attack strategies, enabling the discovery of novel strategies. Our experiments demonstrate AutoMIA can successfully discover new MIAs that are specifically tailored to user-configured target model and dataset, resulting in improvements of up to 0.18 in absolute AUC over existing MIAs. This work provides the first demonstration that LLM agents can serve as an effective and scalable paradigm for designing and implementing MIAs with SOTA performance, opening up new avenues for future exploration.

Tran, Toan Viet [Emory University]↗