Search NASA⌕ Search

SEARCH · Search NASA

Results for “ML T&E”

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 19 records
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

Fixed-point quantum circuits for quantum field theories

Renormalization group ideas and effective operators are introduced to efficiently prepare ground states of massive lattice field theories on digital quantum devices. This is accomplished with a systematic approximation through localized unitaries that removes an exponentially costly barrier in the spatial volume of the quantum simulation. With these methods, classically computed ground states in a spatial volume L, containing a few Compton wavelengths, can be used to determine operators for preparing the ground state toward the thermodynamic limit with a precision improving as e –mL on beyond-classical quantum registers. Here, due to the exponential spatial decay of correlations in massive theories and the double exponential suppression of digitization artifacts in the number of qubits representing the scalar field, the derived fixed-point quantum circuits are expected to be relevant for simulations of quantum field theories throughout the evolution from small-scale near-term quantum devices to large-scale fault-tolerant quantum computers.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗
DOE OSTI2025

ML-driven Strong Lens Discoveries: Down to θ E ~ $0^{_{''}}_{^.}03$ and M halo < 10 11 M ⊙

We present results on extending the strong lens discovery space down to much smaller Einstein radii ( θ E ≲ $0^{_{''}}_{^.}03$) and much lower halo mass (M halo < 10 11 M ⊙ ) through the combination of JWST observations and machine learning (ML) techniques. First, we forecast detectable strong lenses with JWST using CosmoDC2 as the lens catalog, and a source catalog down to 29th magnitude. By further incorporating the VELA hydrodynamical simulations of high-redshift galaxies, we simulate strong lenses. We train a ResNet on these images, achieving near-100% completeness and purity for “conventional” strong lenses ( θ E ≳ $0^{_{''}}_{^.}05$), applicable to JWST, the Hubble Space Telescope (HST), the Roman Space Telescope, and Euclid VIS. For the first time, we also search for very low halo mass strong lenses (M halo < 10 11 M ⊙ ) in simulations, with θ E << $0^{_{''}}_{^.}05$, down to the best resolution ($0^{_{''}}_{^.}03$) and depth (10,000 s) limits of JWST using ResNet. A U-Net model is employed to pinpoint these small lenses in images, which are otherwise virtually impossible for human detection. Our results indicate that JWST can find ∼17/deg 2 such low-halo-mass lenses, with the locations of ∼1.1/deg 2 of these detectable by the U-Net at ∼100% precision (and ∼7.0/deg 2 at a 99.0% precision). To validate our model for finding “conventional” strong lenses, we apply it to HST images, discovering two new strong lens candidates previously missed by human classifiers in a crowdsourcing project (E. O. Garvin et al. 2022). This study demonstrates the (potentially “superhuman”) advantages of ML combined with current and future space telescopes for detecting conventional, and especially, low-halo-mass strong lenses, which are critical for testing cold dark matter models.

Silver, Ethan [Harvard University, Cambridge, MA (↗
DOE OSTI2026

Development of an ab initio learned model of electron deposition range in deuterium-tritium plasmas through time-dependent density functional theory calculations and machine learning

Accurate hydrodynamic modeling for laser-direct-drive (LDD) inertial-confinement-fusion (ICF) relies on precise calculations of the electron thermal conduction in all target materials. The nonlocal stopping range of electrons in ICF plasmas directly influences thermal conduction; yet, no first principles model exists for the electron mean free path in the conduction-zone regime. This work utilized time-dependent stochastic density-functional theory (TD-sDFT) to calculate the electron stopping power in deuterium-tritium (DT) plasmas at (ρ, T) conditions relevant to the conduction zone and the compressed shell in ICF. Using a combination of our TD-sDFT data and already established analytical models, we developed and trained an artificial neural network to create a global model for the nonlocal electron deposition range, λ E . We compared our machine-learning (ML) based model for λ E to the currently-used modified-Lee-More model in LDD radiation-hydrodynamic codes, such as lilac, and saw an overall decrease in the deposition range. To understand the effects of λ E on LDD ICF implosion dynamics, we implemented the ML-based model into lilac; specifically, we looked at designs consistent with a current experiment on the OMEGA laser and for a newly designed LDD-ICF target for the future OMEGA-Next facility. In both cases, we saw an overall drop in predicted ablation pressure, peak areal density, and neutron yield due to the reduced thermal conduction (smaller λ E ) in DT plasmas. Comparisons with the experiment on OMEGA are also made.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
NASA NTRS1972

Development of a tape transport bacterial detection system

The feasibility of a tape transport chemiluminescence system for bacterial monitoring of regenerated water was demonstrated using a manually operated laboratory breadboard. The principle of detection is based on measuring the increase in chemiluminescence produced by the catalytic action of bacterial porphyrins on a luminol-hydrogen peroxide mixture. Viable organisms are distinguished from nonviable by comparing the signals of incubated and unincubated water samples. Using optimized protocols, sensitivities were obtained with 400 ml suspensions of E. coli and Cl. sporogenes. The sensitivity of the unincubated cycle E. coli (aerobe) was found to be 30 to 35 cells/m1, and that of the Cl. sporogenes (anaerobe) was 1000 to 10,000 cells/m1. The lower sensitivity toward Cl. sporogenes is attributed to several factors, namely the lower cytochrome content, the tendency to sporulate, long lag periods and the lower growth rate of Clostridia in general. The operational procedures used for processing the incubated and unincubated samples involved the following sequence: (1) concentrating the sample by filtration through a membrane filter, (2) washing with Dextrose-Thioglycollate Broth (3) incubating (0 to 4 hrs as required), (4) washing with 4M Urea, and (5) reacting with reagent in front of a photomultiplier tube. The signal output was recorded on a strip chart recorder.

Witz, S.↗
DOE OSTI2021

Evaluating Deception Detection Model Robustness To Linguistic Variation

With the increasing use of automated, machine learning-driven tools and the downstream impact that algorithmic judgements can have, it is critical to develop models that are robust to evolving or manipulated inputs. Evaluating the reliability of multimodal models across linguistic variations to understand model susceptibility to intentional linguistic adversarial attacks as well as natural linguistic variations is essential in this pursuit. We present extensive analysis of model robustness and susceptibility to linguistic variations in the setting of deceptive news detection, a difficult classification task that is an increasingly important problem to solve with the impact of misinformation spread online. We evaluate the effectiveness of incorporating adversarial defense strategies and measure model susceptibility to state-of-the-art adversarial attacks using two types of linguistic attacks — character and word perturbations. We consider two multiclass prediction tasks — a 3-way classification of tweets as trustworthy, propaganda, or disinformation; and a 4-way classification as clickbait, hoax, satire, or conspiracy — and compare the performance of three embeddings that have been state-of-the-art for several NLP tasks — GloVe, ELMo, and BERT — to highlight consistent trends in susceptibility, high confidence misclassifications, and high impact failures. We find that character or mixed ensemble models are the most effective defense mechanisms and that character perturbations are a more effective attack than word perturbations for deception classification.

adversarial evaluation↗
DOE OSTI2022

Quaternized chitosan as a biopolymer sanitizer for leafy vegetables: synthesis, characteristics, and traditional vs. dry nano-aerosol applications

A series of quaternary dimethyl-(alkyl)-ammonium chitosan derivatives (QACs) was synthesized and studied for physicochemical properties and bioactivity. The QACs tended to spontaneously self-assembly into nanoaggregates. Antimicrobial activity was examined in vitro on Gram-negative Escherichia coli (E. coli) and Gram-positive Listeria innocua (L. innocua) bacteria as well as phytopathogenic fungus Botrytis cinerea. The hexyl chain-substituted QAC-6 demonstrated the highest potency causing 3.0- and 4.5-log CFU mL-1 reduction of E. coli and L. innocua, respectively. QAC-6 was tested for antimicrobial activity on stainless steel coupons and fresh spinach leaves. A traditional ‘wet’ application (spray) and dry Engineered Water Nanostructure (EWNS) approach were used for spinach decontamination. With both approaches, significant reduction of microbial load on the treated produce was achieved. Finally, the wet application showed a greater reduction of microbial load, while the advantages of EWNS were reaching the antimicrobial effect with miniscule dose of active agent leaving treated surface visibly dry.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2021

Characterization of Infrequent Samples from the Concentration, Storage, and Transfer Facility: F-Area Pump Pit 3 (FPP-3) Sump Sample: December 2021 Sample

In December 2021, SRR-E sent an ~185 mL sample identified as FPP-3 from a F-Area pump pit sump to SRNL for analysis. The sample was pale yellow in color and contained a small amount of suspended solids visually estimated to be less than 1% by volume. SRNL analysis indicated that the sample contained 4.4 E+05 dpm/mL Cs-137, 6.2 E+05 dpm/mL total beta activity, 9.3 E+02 dpm/mL beta activity following cesium removal, and below detectable levels of alpha activity following cesium removal. In addition, the sample contained 0.015 M free OH- and 947 μg total organic carbon/mL.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
DOE OSTI2021

Information-Theoretic Bounds on Quantum Advantage in Machine Learning

We study the performance of classical and quantum machine learning (ML) models in predicting outcomes of physical experiments. The experiments depend on an input parameter x and involve execution of a (possibly unknown) quantum process E. Our figure of merit is the number of runs of E required to achieve a desired prediction performance. We consider classical ML models that perform a measurement and record the classical outcome after each run of E, and quantum ML models that can access E coherently to acquire quantum data; the classical or quantum data are then used to predict the outcomes of future experiments. We prove that for any input distribution D(x), a classical ML model can provide accurate predictions on average by accessing E a number of times comparable to the optimal quantum ML model. In contrast, for achieving an accurate prediction on all inputs, we prove that the exponential quantum advantage is possible. For example, to predict the expectations of all Pauli observables in an n-qubit system ρ, classical ML models require 2 Ω(n) copies of ρ, but we present a quantum ML model using only O(n) copies. Our results clarify where the quantum advantage is possible and highlight the potential for classical ML models to address challenging quantum problems in physics and chemistry.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
NASA NTRS2014

The Impact of Apollo-Era Microbiology on Human Space Flight

The microbiota of crewmembers and the spacecraft environment contributes significant risk to crew health during space flight missions. NASA reduces microbial risk with various mitigation methods that originated during the Apollo Program and continued to evolve through subsequent programs: Skylab, Shuttle, and International Space Station (ISS). A quarantine of the crew and lunar surface samples, within the Lunar Receiving Laboratory following return from the Moon, was used to prevent contamination with unknown extraterrestrial organisms. The quarantine durations for the crew and lunar samples were 21 days and 50 days, respectively. A series of infections among Apollo crewmembers resulted in a quarantine before launch to limit exposure to infectious organisms. This Health Stabilization Program isolated the crew for 21 days before flight and was effective in reducing crew illness. After the program developed water recovery hardware for Apollo spacecraft, the 1967 National Academy of Science Space Science Board recommended the monitoring of potable water. NASA implemented acceptability limits of 10 colony forming units (CFU) per mL and the absence of viable E. coli, anaerobes, yeasts, and molds in three separate 150 mL aliquots. Microbiological investigations of the crew and spacecraft environment were conducted during the Apollo program, including the Apollo-Soyuz Test Project and Skylab. Subsequent space programs implemented microbial screening of the crew for pathogens and acceptability limits on spacecraft surfaces and air. Microbiology risk mitigation methods have evolved since the Apollo program. NASA cancelled the quarantine of the crew after return from the lunar surface, reduced the duration of the Health Stabilization Program; and implemented acceptability limits for spacecraft surfaces and air. While microbial risks were not a main focus of the early Mercury and Gemini programs, the extended duration of Apollo flights resulted in the increased scrutiny of impact of the space flight environment on crew health. The lessons learned during that era of space flight continue to impact microbiology risk mitigation in space programs today.

Elliott, T. F↗
DOE OSTI2022

Accurate Prediction of Voltage of Battery Electrode Materials Using Attention-Based Graph Neural Networks

Performing first-principles calculations to discover electrodes’ properties in the large chemical space is a challenging task. While machine learning (ML) has been applied to effectively accelerate those discoveries, most of the applied methods ignore the materials’ spatial information and only use predefined features: based only on chemical compositions. Here, we propose two attention-based graph convolutional neural network techniques to learn the average voltage of electrodes. Our proposed methods, which combine both atomic composition and atomic coordinates in 3D-space, improve the accuracy in voltage prediction significantly when compared to composition-based ML models. The first model directly learns the chemical reaction of electrodes and metal ions to predict their average voltage, whereas the second model combines electrodes’ ML predicted formation energy (E form ) to compute their average voltage. Our E form -based model demonstrates improved accuracy in transferability from our subset of learned Li ions to Na ions. Moreover, we predicted the theoretical voltage of 10 Na x MPO 4 F (M = Ti, Cr, Fe, Cu, Mn, Co, and Ni) fluorophosphate battery frameworks, which are unavailable in the Material Project database. It could be shown that we can expect average voltages higher than 3.1 V from those Na battery frameworks except from the NaTiPO 4 F and TiPO 4 F pair of electrodes, which offer an average voltage of 1.32 V.

25 ENERGY STORAGE↗
DOE OSTI2021

Leveraging Community and Author Context to Explain the Performance and Bias of Text-Based Deception Detection Models

Deceptive news posts shared in online communities can be detected with NLP models, and much recent research has focused on the development of such models. In this work, we use characteristics of online communities and authors --- the context of how and where content is posted --- to explain the performance of a neural network deception detection model and identify sub-populations who are disproportionately affected by model accuracy or failure. We examine who is posting the content, and where the content is posted to. We find that while author characteristics are better predictors of deceptive content than community characteristics, both characteristics are strongly correlated with model performance. Traditional performance metrics such as F1 score may fail to capture poor model performance on isolated sub-populations such as specific authors, and as such, more nuanced evaluation of deception detection models is critical.

machine learning (ML), machine learning explanatio↗
DOE OSTI2021

Characterization of Infrequent Samples from the Concentration, Storage, and Transfer Facility: H-Area Diversion Box 7 (HDB-7) Sump Sample: December 2021 Sample

In December 2021, SRR-E sent an ~250 mL sample identified as HDB-7 from an H-Area diversion box sump to SRNL for analysis. The sample was clear and colorless and free from any solids. SRNL analysis indicated that the sample contained 2.97E+05 dpm/mL Cs-137, 3.74E+05 dpm/mL total beta activity, 1.04E+03 dpm/mL beta activity following cesium removal, and below detectable levels of alpha activity and alpha activity following cesium removal. In addition, the pH of the sample was 7.32, free hydroxide concentration was 2.09E-07 M, and the density was 0.991 g/mL.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
DOE OSTI2025

A Machine Learning Framework for Predicting Microphysical Properties of Ice Crystals From Cloud Particle Imagery

The microphysical properties of ice crystals are important because they significantly alter the radiative properties and spatiotemporal distributions of clouds, which in turn strongly affect Earth's climate. However, it is challenging to measure key properties of ice crystals, such as mass or morphological features. Here, we present a proof-of-concept framework for predicting three-dimensional (3D) microphysical properties of ice crystals from in situ two-dimensional (2D) imagery. First, we computationally generated synthetic ice crystals using 3D modeling software along with geometric parameters estimated from the 2021 Ice Cryo-Encapsulation Balloon (ICEBall) field campaign. Then, we used synthetic crystals to train machine learning (ML) models to predict effective density ($ρ_e$), effective surface area ($A_e$), and number of bullets ($N_b$) from synthetic rosette imagery. On unseen synthetic images, our ML models accurately predicted ice crystal properties. ResNet-18 performed best, achieving $R^2$ values of 0.99 and 0.98 for $ρ_e$ and $A_e$, respectively, and MAE of 0.10 for mathematical equation in single view tasks. Stereo view ResNet-18 further reduced RMSE by 40% for $ρ_e$ and $A_e$ and reduced MAE by 0.08 for $N_b$. This work provides a novel ML-driven framework for estimating ice microphysical properties from in situ imagery, which will allow for downstream constraints on microphysical parameterizations, such as the mass-size relationship.

Ko, J. [Columbia Univ., New York, NY (United State↗
NASA NTRS2002

Tetracycline rapidly reaches all the constituent cells of uropathogenic Escherichia coli biofilms

We have developed a method for visualizing Escherichia coli cells that are exposed to tetracycline in a biofilm, based on a previous report that liposomes containing the E. coli TetR(B) protein fluoresce when exposed to this antibiotic. By our method, cells devoid of TetR(B) also exhibited tetracycline-dependent fluorescence. At 50 microg of tetracycline ml(-1), planktonic cells of a uropathogenic E. coli (UPEC) strain developed maximal fluorescence after 7.5 to 10 min of exposure. A similar behavior was exhibited by cells in a 24- or 48-h UPEC biofilm, as examined by confocal laser microscopy, regardless of whether they lined empty spaces or occupied densely packed regions. Further, a comparison of phase-contrast and fluorescent images of corresponding biofilm zones showed that all the cells fluoresced. Thus, all the biofilm cells were exposed to tetracycline and there were no pockets within the biofilm where the antibiotic failed to reach. It also appeared unlikely that niches of reduced exposure to the antibiotic existed within the biofilms.

Non-NASA Center↗
DOE OSTI2022

Preliminary Results for Using Uncertainty and Out-of-distribution Detection to Identify Unreliable Predictions.

As machine learning (ML) models are deployed into an ever-diversifying set of application spaces, ranging from self-driving cars to cybersecurity to climate modeling, the need to carefully evaluate model credibility becomes increasingly important. Uncertainty quantification (UQ) provides important information about the ability of a learned model to make sound predictions, often with respect to individual test cases. However, most UQ methods for ML are themselves data-driven and therefore susceptible to the same knowledge gaps as the models themselves. Specifically, UQ helps to identify points near decision boundaries where the models fit the data poorly, yet predictions can score as certain for points that are under-represented by the training data and thus out-of-distribution (OOD). One method for evaluating the quality of both ML models and their associated uncertainty estimates is out-of-distribution detection (OODD). We combine OODD with UQ to provide insights into the reliability of the individual predictions made by an ML model.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1979

Rapid, single-step most-probable-number method for enumerating fecal coliforms in effluents from sewage treatment plants

A single-step most-probable-number method for determining the number of fecal coliform bacteria present in sewage treatment plant effluents is discussed. A single growth medium based on that of Reasoner et al. (1976) and consisting of 5.0 gr. proteose peptone, 3.0 gr. yeast extract, 10.0 gr. lactose, 7.5 gr. NaCl, 0.2 gr. sodium lauryl sulfate, and 0.1 gr. sodium desoxycholate per liter is used. The pH is adjusted to 6.5, and samples are incubated at 44.5 deg C. Bacterial growth is detected either by measuring the increase with time in the electrical impedance ratio between the innoculated sample vial and an uninnoculated reference vial or by visual examination for turbidity. Results obtained by the single-step method for chlorinated and unchlorinated effluent samples are in excellent agreement with those obtained by the standard method. It is suggested that in automated treatment plants impedance ratio data could be automatically matched by computer programs with the appropriate dilution factors and most probable number tables already in the computer memory, with the corresponding result displayed as fecal coliforms per 100 ml of effluent.

Munoz, E. F.↗
NASA NTRS2002

Statistical Properties of Maximum Likelihood Estimators of Power Law Spectra Information

A simple power law model consisting of a single spectral index, a is believed to be an adequate description of the galactic cosmic-ray (GCR) proton flux at energies below 10(exp 13) eV, with a transition at the knee energy, E(sub k), to a steeper spectral index alpha(sub 2) greater than alpha(sub 1) above E(sub k). The Maximum likelihood (ML) procedure was developed for estimating the single parameter alpha(sub 1) of a simple power law energy spectrum and generalized to estimate the three spectral parameters of the broken power law energy spectrum from simulated detector responses and real cosmic-ray data. The statistical properties of the ML estimator were investigated and shown to have the three desirable properties: (P1) consistency (asymptotically unbiased). (P2) efficiency asymptotically attains the Cramer-Rao minimum variance bound), and (P3) asymptotically normally distributed, under a wide range of potential detector response functions. Attainment of these properties necessarily implies that the ML estimation procedure provides the best unbiased estimator possible. While simulation studies can easily determine if a given estimation procedure provides an unbiased estimate of the spectra information, and whether or not the estimator is approximately normally distributed, attainment of the Cramer-Rao bound (CRB) can only he ascertained by calculating the CRB for an assumed energy spectrum-detector response function combination, which can be quite formidable in practice. However. the effort in calculating the CRB is very worthwhile because it provides the necessary means to compare the efficiency of competing estimation techniques and, furthermore, provides a stopping rule in the search for the best unbiased estimator. Consequently, the CRB for both the simple and broken power law energy spectra are derived herein and the conditions under which they are attained in practice are investigated. The ML technique is then extended to estimate spectra information from an arbitrary number of astrophysics data sets produced by vastly different science instruments. This theory and its successful implementation will facilitate the interpretation of spectral information from multiple astrophysics missions and thereby permit the derivation of superior spectral parameter estimates based on the combination of data sets.

Howell, L. W.↗
NASA NTRS2003

Statistical Properties of Maximum Likelihood Estimators of Power Law Spectra Information

A simple power law model consisting of a single spectral index, sigma(sub 2), is believed to be an adequate description of the galactic cosmic-ray (GCR) proton flux at energies below 10(exp 13) eV, with a transition at the knee energy, E(sub k), to a steeper spectral index sigma(sub 2) greater than sigma(sub 1) above E(sub k). The maximum likelihood (ML) procedure was developed for estimating the single parameter sigma(sub 1) of a simple power law energy spectrum and generalized to estimate the three spectral parameters of the broken power law energy spectrum from simulated detector responses and real cosmic-ray data. The statistical properties of the ML estimator were investigated and shown to have the three desirable properties: (Pl) consistency (asymptotically unbiased), (P2) efficiency (asymptotically attains the Cramer-Rao minimum variance bound), and (P3) asymptotically normally distributed, under a wide range of potential detector response functions. Attainment of these properties necessarily implies that the ML estimation procedure provides the best unbiased estimator possible. While simulation studies can easily determine if a given estimation procedure provides an unbiased estimate of the spectra information, and whether or not the estimator is approximately normally distributed, attainment of the Cramer-Rao bound (CRB) can only be ascertained by calculating the CRB for an assumed energy spectrum- detector response function combination, which can be quite formidable in practice. However, the effort in calculating the CRB is very worthwhile because it provides the necessary means to compare the efficiency of competing estimation techniques and, furthermore, provides a stopping rule in the search for the best unbiased estimator. Consequently, the CRB for both the simple and broken power law energy spectra are derived herein and the conditions under which they are stained in practice are investigated.

Howell, L. W., Jr.↗