Search NASA⌕ Search

SEARCH · Search NASA

Results for “Model 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 451 records · Page 25

Field evaluation of semi‐automated moisture estimation from geophysics using machine learning

Geophysical methods can provide three-dimensional (3D), spatially continuous estimates of soil moisture. However, point-to-point comparisons of geophysical properties to measure soil moisture data are frequently unsatisfactory, resulting in geophysics being used for qualitative purposes only. This is because (1) geophysics requires models that relate geophysical signals to soil moisture, (2) geophysical methods have potential uncertainties resulting from smoothing and artifacts introduced from processing and inversion, and (3) results from multiple geophysical methods are not easily combined within a single soil moisture estimation framework. To investigate these potential limitations, an irrigation experiment was performed wherein soil moisture was monitored through time, and several surface geophysical datasets indirectly sensitive to soil moisture were collected before and after irrigation: ground penetrating radar, electrical resistivity tomography (ERT), and frequency domain electromagnetics (FDEM). Data were exported in both raw and processed form, and then snapped to a common 3D grid to facilitate moisture prediction by standard calibration techniques, multivariate regression, and machine learning. A combination of inverted ERT data, raw FDEM, and inverted FDEM data was most informative for predicting soil moisture using a random regression forest model (one-thousand 60/40 training/test cross-validation folds produced root mean squared errors ranging from 0.025–0.046 cm 3 /cm 3 ). This cross-validated model was further supported by a separate evaluation using a test set from a physically separate portion of the study area. Machine learning was conducive to a semi-automated model-selection process that could be used for other sites and datasets to locally improve accuracy.

54 ENVIRONMENTAL SCIENCES↗

Developments of the Jülich–Bonn Dynamical Coupled-Channel Analysis

Here, a new analysis of KΛ photoproduction is presented, including the latest data measured at Jefferson Lab and other facilities. Evidence for two states has been found, the N (1900)3/2 + and N (2060)5/2¯ resonances that have been seen in other analysis. Future extensions are discussed like the use of model selection techniques to achieve a minimal resonance content of the amplitude. Also, conceptual aspects like three-body unitarity are highlighted that can help to provide a connection between phenomenology and ab-initio lattice QCD calculations.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Distributed out-of-memory NMF on CPU/GPU architectures

We propose an efficient distributed out-of-memory implementation of the non-negative matrix factorization (NMF) algorithm for heterogeneous high-performance-computing systems. The proposed implementation is based on prior work on NMFk, which can perform automatic model selection and extract latent variables and patterns from data. In this work, we extend NMFk by adding support for dense and sparse matrix operation on multi-node, multi-GPU systems. The resulting algorithm is optimized for out-of-memory problems where the memory required to factorize a given matrix is greater than the available GPU memory. Memory complexity is reduced by batching/tiling strategies, and sparse and dense matrix operations are significantly accelerated with GPU cores (or tensor cores when available). Input/output latency associated with batch copies between host and device is hidden using CUDA streams to overlap data transfers and compute asynchronously, and latency associated with collective communications (both intra-node and inter-node) is reduced using optimized NVIDIA Collective Communication Library (NCCL) based communicators. Benchmark results show significant improvement, from 32X to 76x speedup, with the new implementation using GPUs over the CPU-based NMFk. Good weak scaling was demonstrated on up to 4096 multi-GPU cluster nodes with approximately 25,000 GPUs when decomposing a dense 340 Terabyte-size matrix and an 11 Exabyte-size sparse matrix of density 10 -6 .

97 MATHEMATICS AND COMPUTING↗

Multimodal hyperspectral optical microscopy

We describe a unique and convenient approach to multimodal hyperspectral optical microscopy, herein achieved by coupling a portable and transferable hyperspectral imager to various optical microscopes. The experimental and data analysis schemes involved in recording spectrally and spatially resolved fluorescence, dark field, and optical absorption micrographs are illustrated through prototypical measurements targeting selected model systems. Namely, hyperspectral fluorescence micrographs of isolated fluorescent beads are employed to ensure spectral calibration of our detector and to gauge the attainable spatial resolution of our measurements; the recorded images are diffraction-limited. Moreover, spatially over-sampled absorption spectroscopy of a single lipid (18:1 Liss Rhod PE) layer reveals that optical densities on the order of 10-3 may be resolved by spatially averaging the recorded optical signatures. We also briefly illustrate two applications of our setup in the general areas of plasmonics and cell biology. Most notably, we deploy hyperspectral optical absorption microscopy to identify and image algal pigments within a single live Tisochrysis lutea cell. Overall, this work paves the way for multimodal multidimensional spectral imaging measurements spanning the realms of several scientific disciples.

47 OTHER INSTRUMENTATION↗

Inverse design of hypoeutectoid pearlite steel microstructures using a deep learning and genetic algorithm optimization framework

Goal-oriented microstructure design in metallic materials is a challenging task due to complex structure-property relationships. Traditional experimental and computational approaches are time-intensive and economically inefficient, limiting their applicability for large-scale design space exploration. Here, in this work, we propose an end-to-end framework that integrates deep learning models with genetic optimization to design microstructures with targeted mechanical properties. Deep learning models enable accurate forward design, while their integration with genetic optimization enables efficient inverse design within a few hours, compared to days or weeks using conventional finite element simulations. The framework combines experimental characterization and finite element modeling to analyze the influence of microstructural features on the mechanical behavior of hypoeutectoid steels. Data from both experiments and simulations are used to train the deep learning models. To demonstrate its effectiveness, we apply the framework to 0.63% carbon steel with proeutectoid ferrite and pearlite phases, commonly used in industrial applications. In this study, 2D microstructures were used for modeling, selected primarily for computational efficiency and to establish proof of concept. The framework successfully optimizes microstructures for targeted yield strength, ultimate strength, and stress concentration factors while significantly reducing computational time. Beyond hypoeutectoid steels, this scalable framework can be extended to other material systems and integrated with additive manufacturing, offering an efficient approach for accelerating microstructure design for specific engineering applications.

ConvLSTM↗

Do Intermediate Radiation Doses Contribute to Late Rectal Toxicity? An Analysis of Data From Radiation Therapy Oncology Group Protocol 94-06

Purpose: To investigate whether the volumes of rectum exposed to intermediate doses, from 30 to 50 Gy, contribute to the risk of Grade {>=}2 late rectal toxicity among patients with prostate cancer receiving radiotherapy. Methods and Materials: Data from 1009 patients treated on Radiation Therapy Oncology Group protocol 94-06 were analyzed using three approaches. First, the contribution of intermediate doses to a previously published fit of the Lyman-Kutcher-Burman (LKB) normal tissue complication probability (NTCP) model was determined. Next, the extent to which intermediate doses provide additional risk information, after taking the LKB model into account, was investigated. Third, the proportion of rectum receiving doses higher than a threshold, VDose, was computed for doses ranging from 5 to 85 Gy, and a multivariate Cox proportional hazards model was used to determine which of these parameters were significantly associated with time to Grade {>=}2 late rectal toxicity. Results: Doses <60 Gy had no detectable impact on the fit of the LKB model, as expected on the basis of the small estimate of the volume parameter (n = 0.077). Furthermore, there was no detectable difference in late rectal toxicity among cohorts with similar risk estimates from the LKB model but with different volumes of rectum exposed to intermediate doses. The multivariate Cox proportional hazards model selected V75 as the only value of VDose significantly associated with late rectal toxicity. Conclusions: There is no evidence from these data that intermediate doses influence the risk of Grade {>=}2 late rectal toxicity. Instead, the critical doses for this endpoint seem to be {>=}75 Gy. It is hypothesized that cases of Grade {>=}2 late rectal toxicity occurring among patients with V75 less than approximately 12% may be due to a 'background' level of risk, likely due mainly to biological factors.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

Study of natural circulation in a VHTR after a LOFA using different turbulence models

Natural convection currents in the core are anticipated in the event of the failure of the gas circulator in a prismatic gas-cooled very high temperature reactor (VHTR). The paths that the helium coolant takes in forming natural circulation loops and the effective heat transport are of interest. The heated flow in the reactor core is turbulent during normal operating conditions and at the beginning of the LOFA with forced convection, but the flow may significantly be slowed down after the event and laminarized with mixed convection. In the present study, the potential occurrence and effective heat transport of natural circulation are demonstrated using computational fluid dynamic (CFD) calculations with different turbulence models as well as laminar flow. Validations and recommendation on turbulence model selection are conducted. The study concludes that large loop natural convection is formed due to the enhanced turbulence levels by the buoyancy effect and the turbulent regime near the interface of upper plenum and flow channels increases the flow resistance for channel flows entering upper plenum and thus less heat can be removed from the core than the prediction by laminar flow assumption.

99 GENERAL AND MISCELLANEOUS↗

A stochastic flow-capturing model to optimize the location of fast-charging stations with uncertain electric vehicle flows

Here, we develop a model to optimize the location of public fast charging stations for electric vehicles (EVs). A difficulty in planning the placement of charging stations is uncertainty in where EV charging demands appear. For this reason, we use a stochastic flow-capturing location model (SFCLM). A sample-average approximation method and an averaged two-replication procedure are used to solve the problem and estimate the solution quality. We demonstrate the use of the SFCLM using a Central-Ohio based case study. We find that most of the stations built are concentrated around the urban core of the region. As the number of stations built increases, some appear on the outskirts of the region to provide an extended charging network. We find that the sets of optimal charging station locations as a function of the number of stations built are approximately nested. We demonstrate the benefits of the charging-station network in terms of how many EVs are able to complete their daily trips by charging midday—six public charging stations allow at least 60% of EVs that would otherwise not be able to complete their daily tours without the stations to do so. We finally compare the SFCLM to a deterministic model, in which EV flows are set equal to their expected values. We show that if a limited number of charging stations are to be built, the SFCLM outperforms the deterministic model. As the number of stations to be built increases, the SFCLM and deterministic model select very similar station locations.

33 ADVANCED PROPULSION SYSTEMS↗

Subgrid modelling for two-dimensional turbulence using neural networks

Here in this investigation, a data-driven turbulence closure framework is introduced and deployed for the subgrid modelling of Kraichnan turbulence. The novelty of the proposed method lies in the fact that snapshots from high-fidelity numerical data are used to inform artificial neural networks for predicting the turbulence source term through localized grid-resolved information. In particular, our proposed methodology successfully establishes a map between inputs given by stencils of the vorticity and the streamfunction along with information from two well-known eddy-viscosity kernels. Through this we predict the subgrid vorticity forcing in a temporally and spatially dynamic fashion. Our study is both a priori and a posteriori in nature. In the former, we present an extensive hyper-parameter optimization analysis in addition to learning quantification through probability-density-function-based validation of subgrid predictions. In the latter, we analyse the performance of our framework for flow evolution in a classical decaying two-dimensional turbulence test case in the presence of errors related to temporal and spatial discretization. Statistical assessments in the form of angle-averaged kinetic energy spectra demonstrate the promise of the proposed methodology for subgrid quantity inference. In addition, it is also observed that some measure of a posteriori error must be considered during optimal model selection for greater accuracy. The results in this article thus represent a promising development in the formalization of a framework for generation of heuristic-free turbulence closures from data.

42 ENGINEERING↗

Guide for Atomic Force Microscopy Image Analysis To Discriminate Heteroatoms in Aromatic Molecules [A Guide for AFM Image Analysis To Discriminate Heteroatoms in Aromatic Molecules]

Heteroatoms are essential for functional groups in organic structures and are complementary to hydrocarbon molecules in reactivities and properties. However, it is still a challenge to quickly identify heteroatoms with only non-contact atomic force microscopy (nc-AFM) to resolve chemical structures in complex unknown mixtures. This study aimed to understand the effect of elemental types on the contrast of atomic force microscopy (AFM) images using a few selected model heterocycles, including dibenzothiophene (DBT), acridine (ACR), and carbazole (CBZ). We identified several features that can be used to find common heteroatoms (S and N) and discriminated them from carbon atoms (C) using nc-AFM images alone. The mechanism of the atom and bond contrast was studied with image simulations and was found to be mostly correlated to van der Waals radii, but other factors, such as bonding geometry, electron density, and substrate interaction, need to be considered. As a result, this work will allow for rapid identification of these common heteroatoms in petroleum with AFM.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Toward Rapid Actinium-225 Purification via Membrane Adsorbers with Covalently Tethered Diglycolamide Ligands

Extractive diglycolamide (DGA) resins are used in several state-of-the-art techniques for purifying 225 Ac, a promising radiometal for targeted alpha therapy. Unfortunately, separation processes that rely on resins are often limited to slow flow rates, high elution volumes, and long processing times. Membrane adsorbers functionalized with DGA ligands are an alternative separation material that may overcome these challenges. This work presents (1) the synthesis of an aminated tetrahexyldiglycolamide ligand, (2) the covalent tethering of the ligand to electrospun poly(vinylbenzyl chloride) fiber mats, and (3) the adsorption and desorption of La(III) and 225 Ac. Chemical and physical characterization supports the covalent tethering of the ligand to the fiber mat, as well as the preservation of the fiber surface area and porosity after functionalization. Equilibrium adsorption experiments were performed with stable La(III) and radioactive 225 Ac. Trends in affinity are consistent between commercial resins and the synthesized membrane adsorbers; however, the Langmuir constants and the maximum binding capacity of the membrane adsorbers were generally lower than the resins. Despite these differences, the modeled selectivity for an equimolar solution of La(III)/ 225 Ac in 10 M nitric acid is 57. Furthermore, 225 Ac is rapidly desorbed from the fibers in 10 M nitric acid (<20 min). The La(III)/ 225 Ac selectivity and rapid 225 Ac desorption indicate this class of materials is promising for rapid radioanalytical separations.

07 ISOTOPE AND RADIATION SOURCES↗

Forecasting the Decay of Hybrid Perovskite Performance Using Optical Transmittance or Reflected Dark-Field Imaging

The practicality and economic viability of hybrid perovskite solar cells hinge on their operational lifetime, and methods for forecasting the performance of perovskites under different operational stresses are urgently needed. Here, we explore the evolution of material-level optoelectronic properties as MAPbI 3 degrades and discover universal behaviors where the carrier diffusion length (L D ) decays before quasi-Fermi-level splitting (ΔE F ), regardless of the specific stress protocol (oxygen, humidity, thermal stress, or a combination). We employ a machine learning greedy feature selection model that uses initially measured properties to predict the time it takes L D to decrease to 85% of its initial value with a prediction accuracy of 12.8%. This model reveals a strong correlation between the initial rate of transmittance change and the time until loss of transport. We translate this material-level finding to photovoltaic device-level forecasting by demonstrating that the rate of change of transmittance is equivalent to the rate of change of the spatial standard deviation of dark-field image intensity (i.e., scattered light intensity) collected in reflection mode (and thus applicable to devices with opaque contacts). Furthermore, this work demonstrates that transmittance and scattering methods are highly effective for accelerated material (and device) stability evaluation and forecasting.

14 SOLAR ENERGY↗

All-Perovskite Multicomponent Nanocrystal Superlattices

Nanocrystal superlattices (NC SLs) have long been sought as promising metamaterials, with nanoscale-engineered properties arising from collective and synergistic effects among the constituent building blocks. Lead halide perovskite (LHP) NCs come across as outstanding candidates for SL design, as they demonstrate collective light emission, known as superfluorescence, in single- and multicomponent SLs. Thus far, LHP NCs have only been assembled in single-component SLs or coassembled with dielectric NC building blocks acting solely as spacers between luminescent NCs. Here, we report the formation of multicomponent LHP NC-only SLs, i.e., using only CsPbBr 3 NCs of different sizes as building blocks. The structural diversity of the obtained SLs encompasses the ABO 6 , ABO 3 , and NaCl structure types, all of which contain orientationally and positionally locked NCs. For the selected model system, the ABO 6 -type SL, we observed efficient NC coupling and Förster-like energy transfer from strongly confined 5.3 nm CsPbBr 3 NCs to weakly confined 17.6 nm CsPbBr 3 NCs, along with characteristic superfluorescence features at cryogenic temperatures. Spatiotemporal exciton dynamics measurements reveal that binary SLs exhibit enhanced exciton diffusivity compared to single-component NC assemblies across the entire temperature range (from 5 to 298 K). The observed coherent and incoherent NC coupling and controllable excitonic transport within the solid NC SLs hold promise for applications in quantum optoelectronic devices.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Homodimerization of the p51 Subunit of HIV-1 Reverse Transcriptase

The dimerization of HIV reverse transcriptase (RT), required to obtain the active form of the enzyme, is influenced by mutations, non-nucleoside reverse transcriptase inhibitors (NNRTIs), nucleotide substrates, Mg ions, temperature, and specifically designed dimerization inhibitors. In this study, we have utilized nuclear magnetic resonance (NMR) spectroscopy of the [methyl-{sup 13}C]methionine-labeled enzyme and small-angle X-ray scattering (SAXS) to investigate how several of these factors influence the dimerization behavior of the p51 subunit. The {sup 1}H-{sup 13}C HSQC spectrum of p51 obtained at micromolar concentrations indicates that a significant fraction of the p51 adopts a 'p66-like' conformation. SAXS data obtained for p51 samples were used to determine the fractions of monomer and dimer in the sample and to evaluate the conformation of the fingers/thumb subdomain. All of the p51 monomer observed was found to adopt the compact, 'p51C' conformation observed for the p51 subunit in the RT heterodimer. The NMR and SAXS data indicate that the p51 homodimer adopts a structure that is similar to the p66/p51 heterodimer, with one p51C subunit and a second p51 subunit in an extended, 'p51E' conformation that resembles the p66 subunit of the heterodimer. The fractional dimer concentration and the fingers/thumb orientation are found to depend strongly on the experimental conditions and exhibit a qualitative dependence on nevirapine and ionic strength (KCl) that is similar to the behavior reported for the heterodimer and the p66 homodimer. The L289K mutation interferes with p51 homodimer formation as it does with formation of the heterodimer, despite its location far from the dimer interface. This effect is readily interpreted in terms of a conformational selection model, in which p51{sub L289K} has a much greater preference for the compact, p51C conformation. A reduced level of dimer formation then results from the reduced ratio of the p51E{sub L289K} to p51C{sub L289K} monomers.

59 BASIC BIOLOGICAL SCIENCES↗

Excited-State Electronic Properties in Zr-Based Metal–Organic Frameworks as a Function of a Topological Network

Molecular assemblies in metal-organic frameworks (MOFs) are reminiscent of natural light-harvesting (LH) systems and considered as emerging materials for energy conversion. Such applications require understanding the correlation between their excited-state properties and underlying topological net. Two chemically identical but topologically different tetraphenylpyrene (1,3,6,8-tetrakis(p-benzoicacid)pyrene; H4TBAPy)-based ZrIV MOFs, NU-901 (scu) and NU-1000 (csq), are chosen to computationally and spectroscopically interrogate the impact of topological difference on their excited-state electronic structures. Time-dependent density functional theory-computed transition density matrices for selected model compounds reveal that the optically relevant S1, S2, and Sn states are delocalized over more than four TBAPy linkers with a maximum exciton size of ~1.7 nm (i.e., two neighboring TBAPy linkers). Computational data further suggests the evolution of polar excitons (hole and electron residing in two different linkers); their oscillator strengths vary with the extent of interchromophoric interaction depending on their topological network. Femtosecond transient absorption (fs-TA) spectroscopic data of NU-901 highlight instantaneous spectral evolution of an intense S1 -> Sn transition at 750 nm, which diminishes with the emergence of a broad (580-1100 nm) induced absorption originating from a fast excimer formation. Although these ultrafast spectroscopic data reveal the first direct spectral observation of fast excimer formation (t = 2 ps) in MOFs, the fs-TA features seen in NU-901 are clearly absent in NU-1000 and the free H4TBAPy linker. Furthermore, transient and steady-state fluorescence data collected as a function of solvent dielectrics reveal that the emissive states in both MOF samples are electronically nonpolar; however, low-lying polar excited states may get involved in the excited-state decay processes in polar solvents. The present work shows that the topological arrangement of the linkers critically controls the excited-state electronic structures.

36 MATERIALS SCIENCE↗

An updated nuclear-physics and multi-messenger astrophysics framework for binary neutron star mergers

The multi-messenger detection of the gravitational-wave signal GW170817, the corresponding kilonova AT2017gfo and the short gamma-ray burst GRB170817A, as well as the observed afterglow has delivered a scientific breakthrough. For an accurate interpretation of all these different messengers, one requires robust theoretical models that describe the emitted gravitational-wave, the electromagnetic emission, and dense matter reliably. In addition, one needs efficient and accurate computational tools to ensure a correct cross-correlation between the models and the observational data. For this purpose, we have developed the Nuclear-physics and Multi-Messenger Astrophysics framework NMMA. The code allows incorporation of nuclear-physics constraints at low densities as well as X-ray and radio observations of isolated neutron stars. In previous works, the NMMA code has allowed us to constrain the equation of state of supranuclear dense matter, to measure the Hubble constant, and to compare dense-matter physics probed in neutron-star mergers and in heavy-ion collisions, and to classify electromagnetic observations and perform model selection. Here, we show an extension of the NMMA code as a first attempt of analyzing the gravitational-wave signal, the kilonova, and the gamma-ray burst afterglow simultaneously. Incorporating all available information, we estimate the radius of a 1.4M ⊙ neutron star to be $R$ $=$ $11.98^{+0.35}_{-0.04}$ km.

79 ASTRONOMY AND ASTROPHYSICS↗

Leveraging unlabeled SEM datasets with self-supervised learning for enhanced particle segmentation

Scanning Electron Microscopes (SEMs) are widely used in experimental science laboratories, often requiring cumbersome and repetitive user analysis. Automating SEM image analysis processes is highly desirable to address this challenge. In particle sample analysis, Machine Learning (ML) has emerged as the most effective approach for particle segmentation. However, the time-intensive process of manually annotating thousands of SEM images limits the applicability of supervised learning approaches. Self-Supervised Learning (SSL) offers a promising alternative by enabling knowledge extraction from raw, unlabeled data. This study presents a framework for evaluating SSL techniques in SEM image analysis, focusing on novel methods leveraging the ConvNeXtV2 architecture for particle detection. A dataset comprising 25,000 SEM images is curated to benchmark these proposed SSL methods. The results demonstrate that ConvNeXtV2 models, with varying parameter counts, consistently outperform other techniques in particle detection across different length scales, achieving up to a 34% reduction in relative error compared to established SSL methods. Furthermore, an ablation study explores the relationship between dataset size and SSL performance, providing actionable insights for practitioners regarding model selection and resource efficiency. This research advances the integration of SSL into autonomous analysis pipelines and supports its application in accelerating materials science discovery.

Rettenberger, Luca↗

DefectTrack: a deep learning-based multi-object tracking algorithm for quantitative defect analysis of in-situ TEM videos in real-time

Abstract In-situ irradiation transmission electron microscopy (TEM) offers unique insights into the millisecond-timescale post-cascade process, such as the lifetime and thermal stability of defect clusters, vital to the mechanistic understanding of irradiation damage in nuclear materials. Converting in-situ irradiation TEM video data into meaningful information on defect cluster dynamic properties (e.g., lifetime) has become the major technical bottleneck. Here, we present a solution called the DefectTrack , the first dedicated deep learning-based one-shot multi-object tracking (MOT) model capable of tracking cascade-induced defect clusters in in-situ TEM videos in real-time. DefectTrack has achieved a Multi-Object Tracking Accuracy (MOTA) of 66.43% and a Mostly Tracked (MT) of 67.81% on the test set, which are comparable to state-of-the-art MOT algorithms. We discuss the MOT framework, model selection, training, and evaluation strategies for in-situ TEM applications. Further, we compare the DefectTrack with four human experts in quantifying defect cluster lifetime distributions using statistical tests and discuss the relationship between the material science domain metrics and MOT metrics. Our statistical evaluations on the defect lifetime distribution suggest that the DefectTrack outperforms human experts in accuracy and speed.

42 ENGINEERING↗