Search NASA⌕ Search

SEARCH · Search NASA

Results for “deep”

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 163 records · Page 9
All sourcesNASA NTRSDOE OSTI
DOE OSTI2019

Deep learning: Extrapolation tool for ab initio nuclear theory

Ab initio approaches in nuclear theory, such as the no-core shell model (NCSM), have been developed for approximately solving finite nuclei with realistic strong interactions. The NCSM and other approaches require an extrapolation of the results obtained in a finite basis space to the infinite basis space limit and assessment of the uncertainty of those extrapolations. Each observable requires a separate extrapolation and many observables have no proven extrapolation method. Herein, we propose a feed-forward artificial neural network (ANN) method as an extrapolation tool to obtain the ground-state energy and the ground-state point-proton root-mean-square (rms) radius along with their extrapolation uncertainties. The designed ANNs are sufficient to produce results for these two very different observables in 6 Li from the ab initio NCSM results in small basis spaces that satisfy the following theoretical physics condition: independence of basis space parameters in the limit of extremely large matrices. Comparisons of the ANN findings with other extrapolation methods are also provided.

6Li↗
DOE OSTI2019

Extraction of Tumor Site from Cancer Pathology Reports using Deep Filters

Purpose: Pathology reports are the primary source of information concerning the millions of cancer cases across the United States. Cancer registries manually process the pathology reports to extract the pertinent information including primary tumor site, behavior, histology, laterality, and grade. Processing a large volume of the pathology reports in a timely manner is a continuing challenge for cancer registries. The purpose of this study is to develop an information extraction pipeline to reliably and efficiently extract reportable information. Method: We have developed a novel inverse-regression (IR) based information extraction pipeline. The inverse-regression based supervised filter has been successfully applied to many application domains. However, its application to the information extraction from unstructured text is hindered primarily by the extreme high-dimensionality of n-gram representations of text. In this study, we attempt to overcome the obstacles by a novel bootstrapping strategy. First, we use an information-theoretic mutual information based filter to discard the excessive and redundant n-gram features. This step reduces the size and improves the condition number of the sample covariance matrix, thus reducing the computational cost and improving the numerical stability of the subsequent inverse-regression step. Then we use localized sliced inverse-regression (LSIR) to learn a low-dimensional discriminatory subspace for information inference. In particular, we use the k-nearest neighbors of an unlabeled pathology report in the learned representation to infer the desired information from the labeled data in a supervised manner. Results: The experiments were conducted on a set of de-identified pathology reports with human expert labels as the ground truth. Our pipeline consistently performed better than or comparable to the best performing state-of-the-art methods while reducing the training and inference times substantially. Conclusion: Our results demonstrate the potential of inverse-regression based information extraction pipeline for reliable and efficient information extraction from unstructured text. The information extracted from the pathology reports can be used along with clinical information, medical imaging, and genomic information to instigate discoveries in cancer research.

Dubey, Abhishek↗
DOE OSTI2016

Deep Vadose Zone Respiration Contributions to Carbon Dioxide Fluxes from a Semiarid Floodplain

Core Ideas A significant fraction of CO 2 fluxes from a semiarid floodplain originates from below the root zone. Measured and calculated field CO 2 fluxes are consistent. Laboratory‐measured respiration rates are consistent with field results. Although CO 2 fluxes from soils are often assumed to originate within shallow soil horizons (<1‐m depth), relatively little is known about respiration rates at greater depths. We compared measured and calculated CO 2 fluxes at the Rifle floodplain along the Colorado River and measured CO 2 production rates of floodplain sediments to determine the relative importance of deeper vadose zone respiration. Calculations based on measured CO 2 gradients and estimated effective diffusion coefficients yielded fluxes that are generally consistent with measurements obtained at the soil surface (326 g C m −2 yr −1 ). Carbon dioxide production from the 2.0‐ to 3.5‐m depth interval was calculated to contribute 17% of the total floodplain respiration, with rates that were larger than some parts of the shallower vadose zone and underlying aquifer. Microbial respiration rates determined from laboratory incubation tests of the sediments support this conclusion. The deeper unsaturated zone typically maintains intermediate water and air saturations, lacks extreme temperatures and salinities, and is annually resupplied with organic carbon from snowmelt‐driven recharge and by water table decline. This combination of favorable conditions supports deeper unsaturated zone microbial respiration throughout the year.

Tokunaga, Tetsu K.↗
DOE OSTI2016

Deep morphological and spectral study of the SNR RCW 86 with Fermi-LAT

RCW 86 is a young supernova remnant (SNR) showing a shell-type structure at several wavelengths and is thought to be an efficient cosmic-ray (CR) accelerator. Earlier Fermi Large Area Telescope results reported the detection of γ-ray emission coincident with the position of RCW 86 but its origin (leptonic or hadronic) remained unclear due to the poor statistics. Thanks to 6.5 years of data acquired by the Fermi-LAT and the new event reconstruction Pass 8, we report the significant detection of spatially extended emission coming from RCW 86. The spectrum is described by a power-law function with a very hard photon index (Γ=1.42±0.1{sub stat}±0.06{sub syst}) in the 0.1–500 GeV range and an energy flux above 100 MeV of (2.91±0.8{sub stat}±0.12{sub syst}) × 10{sup −11} erg cm{sup −2} s{sup −1}. Gathering all the available multiwavelength (MWL) data, we perform a broadband modeling of the nonthermal emission of RCW 86 to constrain parameters of the nearby medium and bring new hints about the origin of the γ-ray emission. For the whole SNR, the modeling favors a leptonic scenario in the framework of a two-zone model with an average magnetic field of 10.2 ± 0.7 μG and a limit on the maximum energy injected into protons of 2 × 10{sup 49} erg for a density of 1 cm{sup −3}. In addition, parameter values are derived for the north–east and south–west (SW) regions of RCW 86, providing the first indication of a higher magnetic field in the SW region.

79 ASTRONOMY AND ASTROPHYSICS↗