Search NASA⌕ Search

SEARCH · Search NASA

Results for “Selective classification”

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 577 records · Page 32

An Iterative Approach to the Feature Selection Problem

The problem dealt with concerns feature selection or reducing the dimension of the data to be processed from n to k. By reducing the dimension of the data from n to k, classification time is generally reduced. Yet the dimension reduction should not be so great that classification accuracy is impaired. Thus, the general problem is considered of classifying an n-dimensional observation vector x into one of m-distinct classes where each class is normally distributed with mean and covariance. It is shown that the probability of misclassification is minimized if a maximum likelihood classification procedure is used to classify the data. The dimension of each observation vector to be processed is conveniently reduced by performing the transformation y = Bx, where B is a K by n matrix of rank k. Thus, the n-dimensional classification problem transforms into a k-dimensional classification problem.

Decell, H. P., Jr.↗

Study of the utilization of EREP data from the Wabash River Basin

The author has identified the following significant results. Analysis of the digitized SL/4 S190A color IR photography proved very difficult. An area within Allen County, including Ft. Wayne, was studied. Eight segments of the study area were clustered separately and the cluster maps were then compared with the photography and maps available. The training areas for the land use classes were selected from the cluster maps. The separability measures (transformed divergence) of the classes indicated that many of the land use classes were not spectrally separable. The classification results bore this out. Visually the resulting classification map was poor, with 67 percent correct data. These results were of significantly lower quality than those obtained for the summertime SL/2 data near Lake Monroe, Indiana. Low contrast between land use classes during the wintertime and the limited spectral range and resolution are the major causes for the poor performance.

Silva, L. F.↗

A search for sulfide-bearing areas using LANDSAT-1 data and digital image-processing techniques

Two experiments were conducted using LANDSAT-1 multispectral scanner data to identify ground features that are indicators of sulfide-bearing rock. A nearby known sulfide deposit was used as a control. In the first experiment, false-color composites were examined visually to locate possible favorable areas; in the second, favorable areas were classified by digital-computer processing. The results of the experiment show that outcrops of hydrothermally altered and mineralized rock can be identified from LANDSAT-1 data under favorable conditions. The empirical method of digital-computer classification of the multispectral scanner data was relatively unrefined and rapid. The five mineralized prospecting sites identified are in locations that the geologist would not have selected as favorable on the basis of geologic knowledge or photogeologic interpretation at the time of the investigation.

Schmidt, R. G.↗

THE HETDEX PILOT SURVEY. I. SURVEY DESIGN, PERFORMANCE, AND CATALOG OF EMISSION-LINE GALAXIES

We present a catalog of emission-line galaxies selected solely by their emission-line fluxes using a wide-field integral field spectrograph. This work is partially motivated as a pilot survey for the upcoming Hobby-Eberly Telescope Dark Energy Experiment. We describe the observations, reductions, detections, redshift classifications, line fluxes, and counterpart information for 397 emission-line galaxies detected over 169 {open_square}' with a 3500-5800 A bandpass under 5 A full-width-half-maximum (FWHM) spectral resolution. The survey's best sensitivity for unresolved objects under photometric conditions is between 4 and 20x 10{sup -17} erg s{sup -1} cm{sup -2} depending on the wavelength, and Ly{alpha} luminosities between 3 x 10{sup 42} and 6 x 10{sup 42} erg s{sup -1} are detectable. This survey method complements narrowband and color-selection techniques in the search of high-redshift galaxies with its different selection properties and large volume probed. The four survey fields within the COSMOS, GOODS-N, MUNICS, and XMM-LSS areas are rich with existing, complementary data. We find 105 galaxies via their high-redshift Ly{alpha} emission at 1.9 < z < 3.8, and the majority of the remainder objects are low-redshift [O II]3727 emitters at z < 0.56. The classification between low- and high-redshift objects depends on rest-frame equivalent width (EW), as well as other indicators, where available. Based on matches to X-ray catalogs, the active galactic nuclei fraction among the Ly{alpha} emitters is 6%. We also analyze the survey's completeness and contamination properties through simulations. We find five high-z, highly significant, resolved objects with FWHM sizes >44 {open_square}' which appear to be extended Ly{alpha} nebulae. We also find three high-z objects with rest-frame Ly{alpha} EW above the level believed to be achievable with normal star formation, EW{sub 0}>240 A. Future papers will investigate the physical properties of this sample.

79 ASTRONOMY AND ASTROPHYSICS↗

Improving the NOvA 3-Flavour Neutrino Oscillation Analysis

NOvA is a long-baseline neutrino experiment studying neutrino oscillations, a quantum mechanical interference phenomena where the observed neutrino flavour differs from that measured earlier, stemming from neutrino mass and flavour states mixing. NOvA consists of two functionally identical tracking calorimeter detectors deployed in the Fermilab NuMI beam. Both detectors are placed 14.6 mrad off the beam axis to achieve a narrow energy peak at 1.8 GeV at an oscillation maximum. The NOvA 3-flavour oscillation analysis measures the neutrino oscillation parameters sin2θ23 and ∆m232 as well as sets limit to δCP , the octant of θ23 and the sign of ∆m32. The event selection for 3-flavour neutrino oscillation analysis ensures the maximum quantity of signal is made available and a minimal amount of background is present. The current disappearance analysis selection has an efficiency of 80% for selecting νμ CC events, and with some improvements additional events could be recovered into the analysis to improve the sensitivity to the aforementioned oscillation parameters. In an effort to recover the currently rejected signal event to the analysis, these events were trained in a classification neural network. The aim of the network was to divide the data into signal (νμ CC events) and background (NC and νe CC events). The highest performing network gave an additional figure of merit gain of 2.34 increasing the sensitivity by 4.3% in effective POT equal to 33 days of additional data taking. This was compared to changing the current particle identification event selection cuts, the best result out of the tested cut combinations gave an additional FOM of 3.36 equal to 3.7% increase in effective POT equivalent to 28 days of data taking.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

A risk index for multicriterial selection of a logging system with low environmental impact

Setting up the working stages in forest operations is conditioned by environmental protection and forest health requirements. This paper exposes a method for improving the decision-making process by choosing the most environmentally effective logging systems according to terrain configuration and stand characteristics. Such a methodology for selecting machines or logging systems accounting for environment, safety as well as economics, becomes mandatory in the context of sustainable management of forest with multiple functions. Based on analytic hierarchy process analysis the following classification of the environmental performance for four considered alternatives was obtained: skyline system (42.43%), forwarder system (20.22%), skidder system (19.92%) and horse logging system (17.43%). Further, an environmental risk matrix for the most important 28 risk factors specific to any work equipment used in forest operations was produced. In the end, a multicriterial analysis generated a risk index RI ranging between 1.0 and 3.5, which could help choosing the optimal combination of logging system and logging equipment with low environmental impact. In order to demonstrate the usefulness of the proposed approach, a simple application in specific conditions of a harvesting site is presented. - Highlights: • We propose a decision-making algorithm to select eco-friendly logging systems. • Analytic hierarchy process was applied for ranking 4 types of logging systems. • An environmental risk matrix with 28 risk factors in forest operations was made up.

54 ENVIRONMENTAL SCIENCES↗

ON THE NATURE OF HARD X-RAY EXTRAGALACTIC SOURCES OBSERVED WITH XMM-NEWTON

Over the last decade, X-ray surveys have provided outstanding new results due to the lack of the common selection effects present at other wavelengths. Here, we have selected a sample of unidentified sources from the XMM-Newton Slew Survey Catalog, likely to be extragalactic. Five of them were observed with the XMM-Newton observatory. In this work, we present the results of the spectral analysis of these objects in the X-ray and optical bands. Only three of them had useful spectroscopic X-ray data, and follow up observations were carried out in the optical range to determine their coordinates, classification, and redshift. The sources are different types of active galactic nuclei (AGNs) with redshifts ranging from 0.059 to 0.386. The properties at both spectral ranges (X-rays and optical) are compatible with the common properties of their types of AGNs. Although the sources were selected by their hard X-ray properties, none of the three detected objects turned out to be an obscured AGN.

79 ASTRONOMY AND ASTROPHYSICS↗

GOODS-HERSCHEL AND CANDELS: THE MORPHOLOGIES OF ULTRALUMINOUS INFRARED GALAXIES AT z {approx} 2

Using deep 100 and 160 {mu}m observations in GOODS-South from GOODS-Herschel, combined with high-resolution HST/WFC3 near-infrared imaging from CANDELS, we present the first detailed morphological analysis of a complete, far-infrared (FIR) selected sample of 52 ultraluminous infrared galaxies (ULIRGs; L{sub IR} > 10{sup 12} L{sub Sun }) at z {approx} 2. We also make use of a comparison sample of galaxies with lower IR luminosities but with the same redshift and H-band magnitude distribution. Our visual classifications of these two samples indicate that the fractions of objects with disk and spheroid morphologies are roughly the same but that there are significantly more mergers, interactions, and irregular galaxies among the ULIRGs (72{sup +5}{sub -7}% versus 32 {+-} 3%). The combination of disk and irregular/interacting morphologies suggests that early-stage interactions, minor mergers, and disk instabilities could play an important role in ULIRGs at z {approx} 2. We compare these fractions with those of a z {approx} 1 sample selected from GOODS-H and COSMOS across a wide luminosity range and find that the fraction of disks decreases systematically with L{sub IR} while the fraction of mergers and interactions increases, as has been observed locally. At comparable luminosities, the fraction of ULIRGs with various morphological classifications is similar at z {approx} 2 and z {approx} 1, though there are slightly fewer mergers and slightly more disks at higher redshift. We investigate the position of the z {approx} 2 ULIRGs, along with 70 z {approx} 2 LIRGs, on the specific star formation rate versus redshift plane, and find 52 systems to be starbursts (i.e., they lie more than a factor of three above the main-sequence relation). We find that many of these systems are clear interactions and mergers ({approx}50%) compared to only 24% of systems on the main sequence relation. If irregular disks are included as potential minor mergers, then we find that up to {approx}73% of starbursts are involved in a merger or interaction at some level. Although the final coalescence of a major merger may not be required for the high luminosities of ULIRGs at z {approx} 2 as is the case locally, the large fraction (50%-73%) of interactions at all stages and potential minor mergers suggests that these processes contribute significantly to the high star formation rates of ULIRGs at z {approx} 2.

79 ASTRONOMY AND ASTROPHYSICS↗

Dust reddened quasars in first and UKIDSS: Beyond the tip of the iceberg

We present the results of a pilot survey to find dust-reddened quasars by matching the Faint Images of the Radio Sky at Twenty-Centimeters (FIRST) radio catalog to the UKIDSS near-infrared survey and using optical data from Sloan Digital Sky Survey to select objects with very red colors. The deep K-band limit provided by UKIDSS allows for finding more heavily reddened quasars at higher redshifts as compared with previous work using FIRST and Two Micron All Sky Survey (2MASS). We selected 87 candidates with K ≤ 17.0 from the UKIDSS Large Area Survey (LAS) First Data Release (DR1), which covers 190 deg{sup 2}. These candidates reach up to ∼1.5 mag below the 2MASS limit and obey the color criteria developed to identify dust-reddened quasars. We have obtained 61 spectroscopic observations in the optical and/or near-infrared, as well as classifications in the literature, and have identified 14 reddened quasars with E(B – V) > 0.1, including 3 at z > 2. We study the infrared properties of the sample using photometry from the Wide-Field Infrared Survey Explorer and find that infrared colors improve the efficiency of red quasar selection, removing many contaminants in an infrared-to-optical color-selected sample alone. The highest-redshift quasars (z ≳ 2) are only moderately reddened, with E(B – V) ∼ 0.2-0.3. We find that the surface density of red quasars rises sharply with faintness, comprising up to 17% of blue quasars at the same apparent K-band flux limit. We estimate that to reach more heavily reddened quasars (i.e., E(B – V) ≳ 0.5) at z > 2 and a depth of K = 17, we would need to survey at least ∼2.5 times more area.

79 ASTRONOMY AND ASTROPHYSICS↗

Bayes classification of terrain cover using normalized polarimetric data

The normalized polarimetric classifier (NPC) which uses only the relative magnitudes and phases of the polarimetric data is proposed for discrimination of terrain elements. The probability density functions (PDFs) of polarimetric data are assumed to have a complex Gaussian distribution, and the marginal PDF of the normalized polarimetric data is derived by adopting the Euclidean norm as the normalization function. The general form of the distance measure for the NPC is also obtained. It is demonstrated that for polarimetric data with an arbitrary PDF, the distance measure of NPC will be independent of the normalization function selected even when the classifier is mistrained. A complex Gaussian distribution is assumed for the polarimetric data consisting of grass and tree regions. The probability of error for the NPC is compared with those of several other single-feature classifiers. The classification error of NPCs is shown to be independent of the normalization function.

Yueh, H. A.↗

A medium-bright quasar sample - New quasar surface densities in the magnitude range from 16.4 to 17.65 for B

A new medium-bright quasar sample (MBQS) is constructed from spectroscopic observations of 140 bright objects selected for varying degrees of blue and ultraviolet excess (B-UVX) in five Palomar 1.2 m Schmidt fields. The MBQS contains 32 quasars with B less than 17.65 mag. The new integral surface densities in the B range from 16.45 to 17.65 mag are approximately 40 percent (or more) higher than expected. The MBQS and its redshift distribution increase the area of the Hubble diagram covered by complete samples of quasars. The general spectroscopic results indicate that the three-color classification process used to catalog the spectroscopic candidates (1) has efficiently separated the intrinsically B-UVX stellar objects from the Population II subdwarfs and (2) has produced samples of B-UVX objects which are more complete than samples selected by (U - B) color alone.

Mitchell, K. J.↗

Comparison of Machine Learning and Deep Learning for View Identification from Cardiac Magnetic Resonance Images

Background: Artificial intelligence is increasingly utilized to aid in the interpretation of cardiac magnetic resonance (CMR) studies. One of the first steps is the identification of the imaging plane depicted, which can be achieved by both deep learning (DL) and classical machine learning (ML) techniques without user input. We aimed to compare the accuracy of ML and DL for CMR view classification and to identify potential pitfalls during training and testing of the algorithms. Methods: To train our DL and ML algorithms, we first established datasets by retrospectively selecting 200 CMR cases. The models were trained using two different cohorts (passively and actively curated) and applied data augmentation to enhance training. Once trained, the models were validated on an external dataset, consisting of 20 cases acquired at another center. We then compared accuracy metrics and applied class activation mapping (CAM) to visualize DL model performance. Results: The DL and ML models trained with the passively-curated CMR cohort were 99.1% and 99.3% accurate on the validation set, respectively. However, when tested on the CMR cases with complex anatomy, both models performed poorly. After training and testing our models again on all 200 cases (active cohort), validation on the external dataset resulted in 95% and 90% accuracy, respectively. The CAM analysis depicted heat maps that demonstrated the importance of carefully curating the datasets to be used for training. Conclusions: Both DL and ML models can accurately classify CMR images, but DL outperformed ML when classifying images with complex heart anatomy.

artificial intelligence↗

Evaluation of trace elements in U.S. coals using the USGS COALQUAL database version 3.0. Part I: Rare earth elements and yttrium (REY)

Coal is a potential source of valuable elements such as rare earth elements and yttrium (REY). In this work, REY concentrations in U.S. domestic coals were evaluated using data from the USGS COALQUAL Database Version 3.0. The database contains a total of 7657 non-weathered, full-bed coal samples. The number of samples containing REY data points varies among elements. Assessment of data quality indicates that some of the REY data are semi-quantitative and should be used with caution. Different analytical instruments and methods with varying accuracies and precisions are thought to be the main sources of errors. Inclusion of qualified data also accounts for the sawtooth pattern of the UCC-normalized REY distribution. A new set of Q factor values was thus proposed to adjust qualified data. Consequently, mean concentrations of REY in U.S. coals were obtained with a total REY concentration of 65.5 ppm on a moisture-free whole coal basis. Further evaluation of REY in 5378 selected coal samples indicates that about 9–13% of the samples fall into the combined category of promising and highly promising coals for REY, according to the classification of Dai et al. (2017). Taking sampling bias into consideration, we further found that bituminous coal, particularly from the central Appalachian region, has the highest probability of being a source for beneficial recovery of REY. More specifically, bituminous coal from eastern Kentucky is likely to be the best option. Lastly, we conclude that U.S. domestic coal is a promising, alternative source for beneficial recovery of REY to meet the U.S. REY demand for economic growth.

01 COAL, LIGNITE, AND PEAT↗

A robust and efficient approach to detect 3D rectal tubes from CT colonography

Purpose: The rectal tube (RT) is a common source of false positives (FPs) in computer-aided detection (CAD) systems for CT colonography. A robust and efficient detection of RT can improve CAD performance by eliminating such ''obvious'' FPs and increase radiologists' confidence in CAD. Methods: In this paper, we present a novel and robust bottom-up approach to detect the RT. Probabilistic models, trained using kernel density estimation on simple low-level features, are employed to rank and select the most likely RT tube candidate on each axial slice. Then, a shape model, robustly estimated using random sample consensus (RANSAC), infers the global RT path from the selected local detections. Subimages around the RT path are projected into a subspace formed from training subimages of the RT. A quadratic discriminant analysis (QDA) provides a classification of a subimage as RT or non-RT based on the projection. Finally, a bottom-top clustering method is proposed to merge the classification predictions together to locate the tip position of the RT. Results: Our method is validated using a diverse database, including data from five hospitals. On a testing data with 21 patients (42 volumes), 99.5% of annotated RT paths have been successfully detected. Evaluated with CAD, 98.4% of FPs caused by the RT have been detected and removed without any loss of sensitivity. Conclusions: The proposed method demonstrates a high detection rate of the RT path, and when tested in a CAD system, reduces FPs caused by the RT without the loss of sensitivity.

62 RADIOLOGY AND NUCLEAR MEDICINE↗

CCS Activities Being Performed by the U.S. DOE

The United States Department of Energy (DOE) is the lead federal agency for the development and deployment of carbon sequestration technologies. Its mission includes promoting scientific and technological innovations and transfer of knowledge for safe and permanent storage of CO 2 in the subsurface. To accomplish its mission, DOE is characterizing and classifying potential geologic storage reservoirs in basins throughout the U.S. and Canada, and developing best practices for project developers, to help ensure the safety of future geologic storage projects. DOE’s Carbon Sequestration Program, Regional Carbon Sequestration Partnership (RCSP) Initiative, administered by the National Energy Technology Laboratory (NETL), is identifying, characterizing, and testing potential injection formations. The RCSP Initiative consists of collaborations among government, industry, universities, and international organizations. Through this collaborative effort, a series of integrated knowledge-based tools have been developed to help potential sequestration project developers. They are the Carbon Sequestration Atlas of the United States and Canada, National Carbon Sequestration Database and Geographic System (NATCARB), and best practice manuals for CCS including Depositional Reservoir Classification for CO 2 ; Public Outreach and Education for Carbon Storage Projects; Monitoring, Verification, and Accounting of CO 2 Stored in Deep Geologic Formation; Site Screening, Site Selection, and Initial Characterization of CO 2 Storage in Deep Geologic Formations. DOE’s future research will help with refinement of these tools and additional best practice manuals (BPM) which focus on other technical aspects of project development.

54 ENVIRONMENTAL SCIENCES↗

Evaluation of Skylab (EREP) data for forest and rangeland surveys

The author has identified the following significant results. Four widely separated sites (near Augusta, Georgia; Lead, South Dakota; Manitou, Colorado; and Redding, California) were selected as typical sites for forest inventory, forest stress, rangeland inventory, and atmospheric and solar measurements, respectively. Results indicated that Skylab S190B color photography is good for classification of Level 1 forest and nonforest land (90 to 95 percent correct) and could be used as a data base for sampling by small and medium scale photography using regression techniques. The accuracy of Level 2 forest and nonforest classes, however, varied from fair to poor. Results of plant community classification tests indicate that both visual and microdensitometric techniques can separate deciduous, conifirous, and grassland classes to the region level in the Ecoclass hierarchical classification system. There was no consistency in classifying tree categories at the series level by visual photointerpretation. The relationship between ground measurements and large scale photo measurements of foliar cover had a correlation coefficient of greater than 0.75. Some of the relationships, however, were site dependent.

Aldrich, R. C.↗

Signal detection theory and methods for evaluating human performance in decision tasks

Signal Detection Theory (SDT) can be used to assess decision making performance in tasks that are not commonly thought of as perceptual. SDT takes into account both the sensitivity and biases in responding when explaining the detection of external events. In the standard SDT tasks, stimuli are selected in order to reveal the sensory capabilities of the observer. SDT can also be used to describe performance when decisions must be made as to the classification of easily and reliably sensed stimuli. Numbers are stimuli that are minimally affected by sensory processing and can belong to meaningful categories that overlap. Multiple studies have shown that the task of categorizing numbers from overlapping normal distributions produces performance predictable by SDT. These findings are particularly interesting in view of the similarity between the task of the categorizing numbers and that of determining the status of a mechanical system based on numerical values that represent sensor readings. Examples of the use of SDT to evaluate performance in decision tasks are reviewed. The methods and assumptions of SDT are shown to be effective in the measurement, evaluation, and prediction of human performance in such tasks.

Obrien, Kevin↗

PSIDD3: Post-Scan Ultrasonic Data Display System for the Windows-Based PC Including Fuzzy Logic Analysis

Post-Scan Interactive Data Display (PSIDD) III is a user-oriented Windows-based system that facilitates the display and comparison of ultrasonic contact data. The system is optimized to compare ultrasonic measurements made at different locations within a material or at different stages of material degradation. PSIDD III provides complete analysis of the primary wave forms in the time and frequency domains along with the calculation of several frequency dependent properties including Phase Velocity and Attenuation Coefficient and several frequency independent properties, like the Cross Correlation Velocity. The system allows image generation on all of the frequency dependent properties at any available frequency (limited by the bandwidth used in the scans) and on any of the frequency independent properties. From ultrasonic contact scans, areas of interest on an image can be studied with regard to underlying raw waveforms and derived ultrasonic properties by simply selecting the point on the image. The system offers various modes of in-depth comparison between scan points. Up to five scan points can be selected for comparative analysis at once. The system was developed with Borland Delphi software (Visual Pascal) and is based on a SQL database. It is ideal for classification of material properties, or location of microstructure variations in materials.

Lovelace, Jeffrey J.↗