Search NASA⌕ Search

SEARCH · Search NASA

Results for “sequential sampling”

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 289 records · Page 16
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

In Situ Diffraction and Ex Situ Transmission X‐Ray Microscopy Studies of Solid‐State Upcycling for NMC Cathodes

Upcycling of recycled LiNi 0.6 Mn 0.2 Co 0.2 O 2 (NMC622) cathodes offers an economical route to produce cathode materials with increased energy density (i.e., LiNi 0.8 Mn 0.1 Co 0.1 O 2 , NMC811) that meet the performance needs of present-day electric vehicles. In this work, solid-state upcycling of NMC622 via calcination with Ni(OH) 2 and LiOH was monitored using in situ synchrotron powder X-ray diffraction measurements. Sequential Rietveld refinements indicate that the calcination proceeds by initially converting Ni(OH) 2 to a rocksalt NiO phase followed by lithiation of NiO to form LiNiO 2 (LNO), with both NMC and LNO phases present in nearly equal proportions at the calcination endpoint. Variable-energy transmission X-ray microscopy tomograms of upcycled samples reveal that the NMC and LNO domains are intermixed at sub-micron length scales. Depth-dependent analysis of multi-elemental fitting maps matches the expected NMC811 composition at the secondary particle level and indicates that transition metal diffusion is not limited by the secondary particle size.

cathode upcycling↗
DOE OSTI2021

Optical properties of differing nanolayered structures of divalent europium doped barium fluoride thin films synthesized by pulsed laser deposition

Optically-active thin films are employed in a variety of applications, such as LEDs and photovoltaics, due to their ability to act as up- or down-photon energy converters. Their performance depends critically on their composition and structure; thus, the use of novel synthesis techniques that allow for their control at the nanoscale level can result in improved efficiency and practicality. Here, layered thin films consisting of Eu 2+ - doped BaF 2 nanocrystalline layers separated by amorphous Al 2 O 3 were synthesized via sequential pulsed laser deposition using three separate targets for the different components; this synthesis technique provides precise control of layer thickness at the nanoscale along with dopant distribution within the film. Cross-sectional transmission electron microscopy analysis verified the desired nano-layering. Post-deposition heat treatments in a nitrogen atmosphere resulted in samples exhibiting steady emission with a broad peak ranging from 400 to 600 nm and a shoulder at 410 nm. The CIE 1931 chromaticity coordinates are x = 0.26-0.29 and y = 0.32-0.35 and vary as a function of the sample configuration. Because the chromaticity coordinates are close to those of a pure white light (x = 0.33, y = 0.33), these films demonstrate advantageous properties for applications with UV-pumped white light LEDs.

36 MATERIALS SCIENCE↗
DOE OSTI2017

Multilevel sequential Monte Carlo: Mean square error bounds under verifiable conditions

We consider the multilevel sequential Monte Carlo (MLSMC) method of Beskos et al. (Stoch. Proc. Appl. [to appear]). This technique is designed to approximate expectations w.r.t. probability laws associated to a discretization. For instance, in the context of inverse problems, where one discretizes the solution of a partial differential equation. The MLSMC approach is especially useful when independent, coupled sampling is not possible. Beskos et al. show that for MLSMC the computational effort to achieve a given error, can be less than independent sampling. In this article we significantly weaken the assumptions of Beskos et al., extending the proofs to non-compact state-spaces. The assumptions are based upon multiplicative drift conditions as in Kontoyiannis and Meyn (Electron. J. Probab. 10 [2005]: 61–123). The assumptions are verified for an example.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2016

Improvement of the thermal stability of nickel silicide using a ruthenium interlayer deposited via remote plasma atomic layer deposition

In this study, the effects of a thin Ru interlayer on the thermal and morphological stability of NiSi have been investigated. Ru and Ni thin films were deposited sequentially to form a Ni/Ru/Si bilayered structure, without breaking the vacuum, by remote plasma atomic layer deposition (RPALD) on a p-type Si wafer. After annealing at various temperatures, the thermal stabilities of the Ni/Ru/Si and Ni/Si structures were investigated by various analysis techniques. The results showed that the sheet resistance of the Ni/Ru/Si sample was consistently lower compared to the Ni/Si sample over the entire temperature range. Although both samples exhibited the formation of NiSi{sub 2} phases at an annealing temperature of 800 °C, as seen with glancing angle x-ray diffraction, the peaks of the Ni/Ru/Si sample were observed to have much weaker intensities than those obtained for the Ni/Si sample. Moreover, the NiSi film with a Ru interlayer exhibited a better interface and improved surface morphologies compared to the NiSi film without a Ru interlayer. These results show that the phase transformation of NiSi to NiSi{sub 2} was retarded and that the smooth NiSi/Si interface was retained due to the activation energy increment for NiSi{sub 2} nucleation that is caused by adding a Ru interlayer. Hence, it can be said that the Ru interlayer deposited by RPALD can be used to control the phase transformation and physical properties of nickel silicide phases.

36 MATERIALS SCIENCE↗
DOE OSTI2015

Simultaneous beam sampling and aperture shape optimization for SPORT

Purpose: Station parameter optimized radiation therapy (SPORT) was recently proposed to fully utilize the technical capability of emerging digital linear accelerators, in which the station parameters of a delivery system, such as aperture shape and weight, couch position/angle, gantry/collimator angle, can be optimized simultaneously. SPORT promises to deliver remarkable radiation dose distributions in an efficient manner, yet there exists no optimization algorithm for its implementation. The purpose of this work is to develop an algorithm to simultaneously optimize the beam sampling and aperture shapes. Methods: The authors build a mathematical model with the fundamental station point parameters as the decision variables. To solve the resulting large-scale optimization problem, the authors devise an effective algorithm by integrating three advanced optimization techniques: column generation, subgradient method, and pattern search. Column generation adds the most beneficial stations sequentially until the plan quality improvement saturates and provides a good starting point for the subsequent optimization. It also adds the new stations during the algorithm if beneficial. For each update resulted from column generation, the subgradient method improves the selected stations locally by reshaping the apertures and updating the beam angles toward a descent subgradient direction. The algorithm continues to improve the selected stations locally and globally by a pattern search algorithm to explore the part of search space not reachable by the subgradient method. By combining these three techniques together, all plausible combinations of station parameters are searched efficiently to yield the optimal solution. Results: A SPORT optimization framework with seamlessly integration of three complementary algorithms, column generation, subgradient method, and pattern search, was established. The proposed technique was applied to two previously treated clinical cases: a head and neck and a prostate case. It significantly improved the target conformality and at the same time critical structure sparing compared with conventional intensity modulated radiation therapy (IMRT). In the head and neck case, for example, the average PTV coverage D99% for two PTVs, cord and brainstem max doses, and right parotid gland mean dose were improved, respectively, by about 7%, 37%, 12%, and 16%. Conclusions: The proposed method automatically determines the number of the stations required to generate a satisfactory plan and optimizes simultaneously the involved station parameters, leading to improved quality of the resultant treatment plans as compared with the conventional IMRT plans.

62 RADIOLOGY AND NUCLEAR MEDICINE↗
DOE OSTI2017

Revisiting the Anionic Polymerization of Methyl Ethacrylate

Synthesis of poly(methyl ethacrylate), (PMEA), in tetrahydrofuran at -78 °C using anionic polymerization techniques results in high molar mass (>30 kg mol-1), low dispersity (1.3), and high conversion (>81%). The molar masses of a series of samples are consistent with values anticipated by the monomer-to-initiator ratio and conversion. These results represent a significant improvement to earlier reported attempts to prepare PMEA using anionic methods. Successful diblock polymerization of polystyrene-block-PMEA, (PS-PMEA), and poly(4-tert-butylstyrene)-block-PMEA, (PtBS-PMEA), is achieved through sequential anionic polymerization techniques with dispersities as low as 1.06 and segment molar fractions close to those targeted. Broad principal scattering peaks observed by small-angle X-ray scattering (SAXS) for symmetric PS-PMEA at relatively high molar mass (39 kg mol-1) suggests an effective interaction parameter (χeff) that is smaller than for PS-block-poly(methyl methacrylate). On the other hand, PtBS-PMEA block polymers form a well-ordered morphology based on SAXS measurements and is attributable to the more hydrophobic PtBS segment. These results confirm the viability of PMEA as a new constituent in the expanding suite of polymers suitable for preparing nanostructured block polymers.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Phosphorus chemistry in plant charcoal: interplay between biomass composition and thermal condition

Background Vegetation fire may change Phosphorus (P) cycling in terrestrial ecosystems through converting biomass into fire residues. Aim The aim of this study was to understand the chemistry and mobility of P in fire residues to help reveal P thermochemistry during biomass burning and post-fire P cycling. Methods A combination of sequential extraction, liquid 31P NMR and P K-edge XANES was used to obtain quantitative P speciation and explain P solubilisation behaviours of charcoal. Key results Despite varying diverse P species existing in raw biomass, only two P structural moieties – orthophosphate and pyrophosphate – were identified in charcoal. However, relative abundance of pyrophosphate differs greatly among charcoal samples from different biomass types, ranging between 0 and 40% of total extractable P. Although P K-edge XANES data indicates abundant soluble phosphate minerals, most P (70–90%) is likely occluded physically in the charcoal. The bicarbonate-extractable P (the Olsen-P) varies significantly and cannot be explained by surface P concentration or elemental stoichiometry alone. Conclusion and implications The results suggest the importance of starting biomass P speciation (i.e. molecular structure and complexation environment) and thermal conditions in controlling P speciation and availability in charcoal. The different P chemistry between charcoal and ash suggests the importance of fire types and severity in disturbing the P cycle.

Forestry↗
DOE OSTI2024

Physics-inspired spatiotemporal-graph AI ensemble for the detection of higher order wave mode signals of spinning binary black hole mergers

We present a new class of AI models for the detection of quasi-circular, spinning, non-precessing binary black hole mergers whose waveforms include the higher order gravitational wave modes ($\ell$, |m|) = {(2,2), (2,1), (3,3), (3,2), (4,4)}, and mode mixing effects in the $\ell$ = 3, |m| = 2 harmonics. These AI models combine hybrid dilated convolution neural networks to accurately model both short- and long-range temporal sequential information of gravitational waves; and graph neural networks to capture spatial correlations among gravitational wave observatories to consistently describe and identify the presence of a signal in a three detector network encompassing the Advanced LIGO and Virgo detectors. We first trained these spatiotemporal-graph AI models using synthetic noise, using 1.2 million modeled waveforms to densely sample this signal manifold, within 1.7 h using 256 NVIDIA A100 GPUs in the Polaris supercomputer at the Argonne Leadership Computing Facility. This distributed training approach exhibited optimal classification performance, and strong scaling up to 512 NVIDIA A100 GPUs. With these AI ensembles we processed data from a three detector network, and found that an ensemble of 4 AI models achieves state-of-the-art performance for signal detection, and reports two misclassifications for every decade of searched data. We distributed AI inference over 128 GPUs in the Polaris supercomputer and 128 nodes in the Theta supercomputer, and completed the processing of a decade of gravitational wave data from a three detector network within 3.5 h. Finally, we fine-tuned these AI ensembles to process the entire month of February 2020, which is part of the O3b LIGO/Virgo observation run, and found 6 gravitational waves, concurrently identified in Advanced LIGO and Advanced Virgo data, and zero false positives. This analysis was completed in one hour using one NVIDIA A100 GPU.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2016

Phosphate-Induced Immobilization of Uranium in Hanford Sediments

Phosphate can be added to subsurface environments to immobilize U(VI) contamination. The efficacy of immobilization depends on the site-specific groundwater chemistry and aquifer sediment properties. Batch and column experiments were performed with sediments from the Hanford 300 Area in Washington State and artificial groundwater prepared to emulate the conditions at the site. Batch experiments revealed enhanced U(VI) sorption with increasing phosphate addition. X-ray absorption spectroscopy measurements of samples from the batch experiments found that U(VI) was predominantly adsorbed at conditions relevant to the column experiments and most field sites (low U(VI) loadings, <25 μM), and U(VI) phosphate precipitation occurred only at high initial U(VI) (>25μM) and phosphate loadings. While batch experiments showed the transition of U(VI) uptake from adsorption to precipitation, the column study was more directly relevant to the subsurface environment because of the high solid:water ratio in the column and the advective flow of water. In column experiments, nearly six times more U(VI) was retained in sediments when phosphate-containing groundwater was introduced to U(VI)-loaded sediments than when the groundwater did not contain phosphate. This enhanced retention persisted for at least one month after cessation of phosphate addition to the influent fluid. Sequential extractions and laser-induced fluorescence spectroscopy of sediments from the columns suggested that the retained U(VI) was primarily in adsorbed forms. These results indicate that in situ remediation of groundwater by phosphate addition provides lasting benefit beyond the treatment period via enhanced U(VI) adsorption to sediments.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2022

Titanium isotope systematics of refractory inclusions: Echoes of molecular cloud heterogeneity

Calcium-aluminum-rich inclusions (CAIs) are highly refractory objects found in different chondrite groups and represent some of the oldest known solids of the Solar System. As such, CAIs provide key information regarding the conditions prevailing in the solar protoplanetary disk as well as subsequent mixing and transport processes. Many studies have investigated CAIs for their isotopic compositions and reported nucleosynthetic isotope anomalies in numerous elements, which are typically explained by the variable incorporation of isotopically highly anomalous presolar phases. However, with the exception of 54 Cr-enriched nanospinels, the exact presolar phases responsible for the isotopic heterogeneities are yet to be identified. To address this issue, we here present in-situ Ti isotopic analyses obtained on a diverse set of CAIs from various CV3 chondrites. In this work, the in-situ measurements were performed by targeting individual mineral phases of 15 CAIs with laser-ablation mass spectrometry and indicate significant inter- and intra-CAI isotopic heterogeneity in the neutron-rich isotope 50 Ti. This is particularly pronounced for primitive fine-grained CAIs, whereas coarse-grained CAIs, which have been subject to melting, exhibit smaller degrees of Ti isotopic heterogeneity. To further investigate this Ti isotopic heterogeneity, we additionally obtained Ti isotopic compositions of sequential acid leachates from two fine-grained and two coarse-grained CAIs derived from CV3 chondrites. In contrast to potential expectations from the first part of the study, we do not observe any significant intra-CAI Ti isotopic heterogeneity between the different leaching steps. The lack of intra-CAI Ti isotopic heterogeneity in the acid leachate samples of this study likely reflects that the leaching procedure is unable to efficiently separate the carriers of isotopically anomalous Ti in CAIs. By comparing the bulk CAI Ti isotope compositions with Ti isotope data for hibonite-rich objects from the literature, we find that the range of Ti isotope compositions recorded by CAIs from various chondrite groups can be accounted for by the averaging of hibonite grains. In turn, the variable Ti isotope compositions of hibonite grains can be explained by the averaging of isotopically diverse presolar grains present in the Sun’s parental molecular cloud. This effect of averaging is statistically supported by the central limit theorem, and the concept has the potential to be useful for other isotopic systems.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

Exabiome: Advancing Microbial Science through Exascale Computing

The Exabiome project seeks to improve the understanding of microbiomes through the development of methods for accelerating metagenomic science using exascale computing. This article gives an overview of scientific impact of the three components of the project: metagenome assembly, protein family detection, and comparative analysis of metagenomes. Exabiome developed MetaHipMer, the only metagenome assembler capable of scaling to full exascale systems. MetaHipMer has enabled ground-breaking assemblies on the Frontier supercomputer, with many scientific benefits, such as the discovery of rare species and viral genomes. To investigate protein families, Exabiome developed two exascale tools, PASTIS and HipMCL. Together, these can utilize exascale resources to understand the functional diversity of billions of dark matter proteins and novel protein families. For comparative analysis, Exabiome developed kmerprof, a tool that can be used to compare huge metagenomes for many different scientific purposes, for example, grouping human microbiomes according to body location.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2021

Tailoring plasticity mechanisms in compositionally graded hierarchical steels fabricated using additive manufacturing

Abstract While there exists in nature abundant examples of materials with site-specific gradients in microstructures and properties, engineers and designers have traditionally used monolithic materials with discrete properties. Now, however, additive manufacturing (AM) offers the possibility of creating structures that mimic some aspects of nature. One example that has attracted attention in the recent years is the hierarchical structure in bamboo. The hierarchical architecture in bamboo is characterized by spatial gradients in properties and microstructures and is well suited to accommodate and survive complex stress states, severe mechanical forces, and large deformations. While AM has been used routinely to fabricate functionally graded materials, this study distinguishes itself by leveraging AM and physical metallurgy concepts to trigger cascading deformation in a single sample. Specifically, we have been successful in using AM to fabricate steel with unique spatial hierarchies in structure and property to emulate the structure and deformation mechanisms in natural materials. This study shows an improvement in the strength and ductility of the nature-inspired “hierarchical steel” compared with conventional cast stainless steels. In situ characterization proves that this improvement is due to the sequential activation of multiple deformation mechanisms namely twinning, transformation-induced plasticity, and dislocation-based plasticity. While significantly higher strengths can be achieved by refining the chemical and processing technique, this study sets the stage to achieve the paradigm of using AM to fabricate structures which emulate the flexibility in mechanical properties of natural materials and are able to adapt to in-service conditions.

36 MATERIALS SCIENCE↗
DOE OSTI2019

Rare earth elements and radiogenic strontium isotopes in carbonate minerals reveal diagenetic influence in shales and limestones in the Appalachian Basin

Rare earth element (REE) signatures are often applied to interpret paleoenvironmental conditions in sedimentary basins, however the complicated diagenetic histories in dynamic depositional environments can affect interpretation of measured REE signatures. Prior studies on REE content in shales and limestones indicated that REE in specific mineral phases may provide unique information concerning diagenetic reactions occurring in sedimentary rocks during burial and compaction. Application of sequential extraction techniques to target REE signatures in specific mineral fractions may provide greater insight into the complex processes that occurred during diagenesis and catagenesis. Thus, using this technique, this study provides a detailed account for REE, 87 Sr/ 86 Sr, and δ 13 C in the carbonate fraction of the Marcellus Shale and adjacent formations in the Appalachian Basin. A suite of 49 rock samples collected from two cores recovered from Monongalia County, West Virginia, USA was analyzed. The results showed that the carbonate concretions in calcareous shales and carbonate cements in black shales were very distinct. The REE plus yttrium (REY) were more concentrated in the carbonate concretions than in the carbonate cements. The carbonate concretions displayed REY patterns that are closely similar to modern seawater, while MREE-enrichment was observed in the carbonate cements. Likewise, the 87 Sr/ 86 Sr values in the carbonate concretions were similar to those measured for unaltered Middle Devonian carbonates, while the 87 Sr/ 86 Sr in the carbonate cements were much more radiogenic. This observation indicates that the carbonate cement could have been inherited radiogenic 87 Sr expulsed from clays during illite-smectite transition. Overall, this study demonstrated two distinct processes involved with controlling the carbonate geochemistry within Appalachian Basin shales and limestones: one fraction displaying minimal diagenetic alteration relative to depositional conditions (carbonate concretions in calcareous shales and limestone carbonate in limestones), and another fraction displaying the evidence for extensive chemical alteration during illite-smectite transition and catagenesis (carbonate cements in black shales). Here, in addition, highly variable and significant enrichments of U and Mo demonstrate that the Union Spring member and the Lower Oatka Creek member of the Marcellus Shale in the southwestern Appalachian Basin (MSEEL site) were deposited under mainly anoxic environment whereas intermitted episodes of dysoxic to perhaps oxic conditions occurred during the deposition of the Upper Oatka Creek member.

58 GEOSCIENCES↗
DOE OSTI2015

Tunability of exchange bias in Ni@NiO core-shell nanoparticles obtained by sequential layer deposition

Films of magnetic Ni@NiO core-shell nanoparticles (NPs, core diameter d congruent to 12 nm, nominal shell thickness variable between 0 and 6.5 nm) obtained with sequential layer deposition were investigated, to gain insight into the relationships between shell thickness/morphology, core-shell interface, and magnetic properties. Different values of NiO shell thickness t(s) could be obtained while keeping the Ni core size fixed, at variance with conventional oxidation procedures where the oxide shell is grown at the expense of the core. Chemical composition, morphology of the as-produced samples and structural features of the Ni/NiO interface were investigated with x-ray photoelectron spectroscopy and microscopy (scanning electron microscopy, transmission electron microscopy) techniques, and related with results from magnetic measurements obtained with a superconducting quantum interference device. The effect of the shell thickness on the magnetic properties could be studied. The exchange bias (EB) field H-bias is small and almost constant for t(s) up to 1.6 nm; then it rapidly grows, with no sign of saturation. This behavior is clearly related to the morphology of the top NiO layer, and is mostly due to the thickness dependence of the NiO anisotropy constant. The ability to tune the EB effect by varying the thickness of the last NiO layer represents a step towards the rational design and synthesis of core-shell NPs with desired magnetic properties.

77 NANOSCIENCE AND NANOTECHNOLOGY↗
DOE OSTI2022

Size-Dependent Nucleation in Crystal Phase Transition from Machine Learning Metadynamics

In this Letter, we present a framework that combines machine learning potential (MLP) and metadynamics to investigate solid-solid phase transition. Here, based on the spectral descriptors and neural networks regression, we develop a scalable MLP model to warrant an accurate interpolation of the energy surface where two phases coexist. Applying it to the simulation of B4–B1 phase transition of GaN under 50 GPa with different model sizes, we observe sequential change of the phase transition mechanism from collective modes to nucleation and growths. When the size is at or below 128 000 atoms, the nucleation and growth appear to follow a preferred direction. At larger sizes, the nuclei occur at multiple sites simultaneously and grow to microstructures by passing the critical size. The observed change of the atomistic mechanism manifests the importance of statistical sampling with large system size in phase transition modeling.

36 MATERIALS SCIENCE↗
DOE OSTI2017

Vacancy effects on the formation of helium and krypton cavities in 3-C-SiC irradiated and annealed at elevated temperatures

Polycrystalline 3C-SiC was sequentially irradiated at 400 and 750°C with 120 keV He 2+ and 4 MeV Kr 15+ ions to 10 17 and 41016 cm -2 , respectively. The Kr 15+ ions penetrated the entire depth of the He 2+ ion implantation region. Three areas of He 2+ , Kr 15+ and He 2+ + Kr 15+ ion implanted 3C-SiC were created through masked overlapping irradiations. The sample was subsequently annealed at 1600°C in vacuum and characterized using cross-sectional transmission electron microscopy and energy-dispersive x-ray spectroscopy. Compared to the He 2+ ion only implanted 3C-SiC, helium cavities in the He 2+ and Kr15+ co-implanted 3C-SiC had a smaller size but higher density. At 25 dpa, presence of He in the co-implanted 3C-SiC significantly promoted cavity growth; much smaller voids were formed in the Kr 15+ ion only irradiated 3C-SiC at the same dose. In addition, local Kr migration and trapping at cavities occurred, but long-range Kr diffusion in 3C-SiC was not observed up to 1600°C.

36 MATERIALS SCIENCE↗
DOE OSTI2015

Fast and Accurate Support Vector Machines on Large Scale Systems

Support Vector Machines (SVM) is a supervised Machine Learning and Data Mining (MLDM) algorithm, which has become ubiquitous largely due to its high accuracy and obliviousness to dimensionality. The objective of SVM is to find an optimal boundary --- also known as hyperplane --- which separates the samples (examples in a dataset) of different classes by a maximum margin. Usually, very few samples contribute to the definition of the boundary. However, existing parallel algorithms use the entire dataset for finding the boundary, which is sub-optimal for performance reasons. In this paper, we propose a novel distributed memory algorithm to eliminate the samples which do not contribute to the boundary definition in SVM. We propose several heuristics, which range from early (aggressive) to late (conservative) elimination of the samples, such that the overall time for generating the boundary is reduced considerably. In a few cases, a sample may be eliminated (shrunk) pre-emptively --- potentially resulting in an incorrect boundary. We propose a scalable approach to synchronize the necessary data structures such that the proposed algorithm maintains its accuracy. We consider the necessary trade-offs of single/multiple synchronization using in-depth time-space complexity analysis. We implement the proposed algorithm using MPI and compare it with libsvm--- de facto sequential SVM software --- which we enhance with OpenMP for multi-core/many-core parallelism. Our proposed approach shows excellent efficiency using up to 4096 processes on several large datasets such as UCI HIGGS Boson dataset and Offending URL dataset.

Vishnu, Abhinav↗
DOE OSTI2022

ADAPTIVE GROUP TESTING WITH MISMATCHED MODELS

Accurate detection of infected individuals is one of the critical steps in stopping any pandemic. When the underlying infection rate of the disease is low, testing people in groups, instead of testing each individual in the population, can be more efficient. In this work, we consider noisy adaptive group testing design with specific test sensitivity and specificity that select the optimal group given previous test results based on pre-selected utility function. As in prior studies on group testing, we model this problem as a sequential Bayesian Optimal Experimental Design (BOED) to adaptively design the groups for each test. We analyze the required number of group tests when using the updated posterior on the infection status and the corresponding Mutual Information (MI) as our utility function for selecting new groups. More importantly, we study how the potential bias on the ground-truth noise of group tests may affect the group testing sample complexity.

97 MATHEMATICS AND COMPUTING↗