Search NASA⌕ Search

SEARCH · Search NASA

Results for “clustering analysis”

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 145 records · Page 8

Use of UV Sources for Detection and Identification of Explosives

Measurement of Raman and native fluorescence emission using ultraviolet (UV) sources (<400 nm) on targeted materials is suitable for both sensitive detection and accurate identification of explosive materials. When the UV emission data are analyzed using a combination of Principal Component Analysis (PCA) and cluster analysis, chemicals and biological samples can be differentiated based on the geometric arrangement of molecules, the number of repeating aromatic rings, associated functional groups (nitrogen, sulfur, hydroxyl, and methyl), microbial life cycles (spores vs. vegetative cells), and the number of conjugated bonds. Explosive materials can be separated from one another as well as from a range of possible background materials, which includes microbes, car doors, motor oil, and fingerprints on car doors, etc. Many explosives are comprised of similar atomic constituents found in potential background samples such as fingerprint oils/skin, motor oil, and soil. This technique is sensitive to chemical bonds between the elements that lead to the discriminating separability between backgrounds and explosive materials.

Hug, William↗

SU-F-R-08: Can Normalization of Brain MRI Texture Features Reduce Scanner-Dependent Effects in Unsupervised Machine Learning?

Purpose: To reduce differences in features calculated from MRI brain scans acquired at different field strengths with or without Gadolinium contrast. Methods: Brain scans were processed for 111 epilepsy patients to extract hippocampus and thalamus features. Scans were acquired on 1.5 T scanners with Gadolinium contrast (group A), 1.5T scanners without Gd (group B), and 3.0 T scanners without Gd (group C). A total of 72 features were extracted. Features were extracted from original scans and from scans where the image pixel values were rescaled to the mean of the hippocampi and thalami values. For each data set, cluster analysis was performed on the raw feature set and for feature sets with normalization (conversion to Z scores). Two methods of normalization were used: The first was over all values of a given feature, and the second by normalizing within the patient group membership. The clustering software was configured to produce 3 clusters. Group fractions in each cluster were calculated. Results: For features calculated from both the non-rescaled and rescaled data, cluster membership was identical for both the non-normalized and normalized data sets. Cluster 1 was comprised entirely of Group A data, Cluster 2 contained data from all three groups, and Cluster 3 contained data from only groups 1 and 2. For the categorically normalized data sets there was a more uniform distribution of group data in the three Clusters. A less pronounced effect was seen in the rescaled image data features. Conclusion: Image Rescaling and feature renormalization can have a significant effect on the results of clustering analysis. These effects are also likely to influence the results of supervised machine learning algorithms. It may be possible to partly remove the influence of scanner field strength and the presence of Gadolinium based contrast in feature extraction for radiomics applications.

60 APPLIED LIFE SCIENCES↗

Scalable, In-situ Data Clustering Data Analysis for Extreme Scale Scientific Computing (Final Report)

The objective of this project is to address challenges in the design and development of scalable in-situ data clustering and analytics algorithms and software. Our goal is to develop parallel software consisting of a set of spatio-temporal data clustering and anomaly detection functions, both of which are very important for large-scale analysis and have wide applicability for in-situ runs as well as post-processing analysis. Our design principles for in-situ analysis consider the following: (1) identify parts of the computation can be done close to the data within the nodes, while it is still in memory; (2) extract analysis components can (and should) be performed in remote staging and analysis nodes; (3) develop error-bound approximation methods for applications tolerable for small errors; (4) identify the type of derived distributions and statistics, for spatio-temporal data, that can be kept locally in order to both accelerate computations and meet energy constraints in subsequent iterations and phases; (5) use a self-describing data format so that data can be consistent and understood among local storage (memory and SSDs) and at staging and analysis nodes, thereby providing portability and flexibility; (6) develop service-oriented functions that can schedule in-situ and post-hoc analysis tasks based on the dynamic requirements of applications. Our development focus is to produce the parallel data analysis software/library that will be scalable, reusable, extensible, and generic for applications in different disciplines. The software will be able to run in-situ with the simulations as well as post-hoc analysis. This approach will satisfy many synergistic requirements for data intensive applications executed on data coming from instruments and experiments. In particular, the proposed multilevel approach is directly applicable to perform design tradeoffs for running part of the algorithms near the instruments and the rest on remote (analysis) systems.

97 MATHEMATICS AND COMPUTING↗

Lumping and splitting: Toward a classification of mineral natural kinds

How does one best subdivide nature into kinds? All classification systems require rules for lumping similar objects into the same category, while splitting differing objects into separate categories. Mineralogical classification systems are no exception. Our work in placing mineral species within their evolutionary contexts necessitates this lumping and splitting because we classify “mineral natural kinds” based on unique combinations of formational environments and continuous temperature-pressure-composition phase space. Consequently, we lump two minerals into a single natural kind only if they: (1) are part of a continuous solid solution; (2) are isostructural or members of a homologous series; and (3) form by the same process. A systematic survey based on these criteria suggests that 2310 (~41%) of 5659 IMA-approved mineral species can be lumped with one or more other mineral species, corresponding to 667 “root mineral kinds,” of which 353 lump pairs of mineral species, while 129 lump three species. Eight mineral groups, including cancrinite, eudialyte, hornblende, jahnsite, labuntsovite, satorite, tetradymite, and tourmaline, are represented by 20 or more lumped IMA-approved mineral species. A list of 5659 IMA-approved mineral species corresponds to 4016 root mineral kinds according to these lumping criteria. The evolutionary system of mineral classification assigns an IMA-approved mineral species to two or more mineral natural kinds under either of two splitting criteria: (1) if it forms in two or more distinct paragenetic environments, or (2) if cluster analysis of the attributes of numerous specimens reveals more than one discrete combination of chemical and physical attributes. A total of 2310 IMA-approved species are known to form by two or more paragenetic processes and thus correspond to multiple mineral natural kinds; however, adequate data resources are not yet in hand to perform cluster analysis on more than a handful of mineral species. We find that 1623 IMA-approved species (~29%) correspond exactly to mineral natural kinds; i.e., they are known from only one paragenetic environment and are not lumped with another species in our evolutionary classification. Greater complexity is associated with 587 IMA-approved species that are both lumped with one or more other species and occur in two or more paragenetic environments. In these instances, identification of mineral natural kinds may involve both lumping and splitting of the corresponding IMA-approved species on the basis of multiple criteria. Based on the numbers of root mineral kinds, their known varied modes of formation, and predictions of minerals that occur on Earth but are as yet undiscovered and described, we estimate that Earth holds more than 10000 mineral natural kinds.

Philosophy of mineralogy↗

Identification of a typical flight patterns

Method and system for analyzing aircraft data, including multiple selected flight parameters for a selected phase of a selected flight, and for determining when the selected phase of the selected flight is atypical, when compared with corresponding data for the same phase for other similar flights. A flight signature is computed using continuous- valued and discrete-valued flight parameters for the selected flight parameters and is optionally compared with a statistical distribution of other observed flight signatures, yielding a typicality scores for the same phase for other similar flights. A cluster analysis is optionally applied to the flight signatures to define an optimal collection of clusters. A level of atypicality for a selected flight is estimated, based upon an index associated with the cluster analysis.

Irving C Statler↗

Identification of atypical flight patterns

Method and system for analyzing aircraft data, including multiple selected flight parameters for a selected phase of a selected flight, and for determining when the selected phase of the selected flight is atypical, when compared with corresponding data for the same phase for other similar flights. A flight signature is computed using continuous-valued and discrete-valued flight parameters for the selected flight parameters and is optionally compared with a statistical distribution of other observed flight signatures, yielding atypicality scores for the same phase for other similar flights. A cluster analysis is optionally applied to the flight signatures to define an optimal collection of clusters. A level of atypicality for a selected flight is estimated, based upon an index associated with the cluster analysis.

Statler, Irving C.↗

Cluster Expansion Analysis of Atomic Order in Li-Ion Battery Cathode Material LiCo y Ni 1-y O 2

A modified cluster-expansion treatment was developed recently to analyze atomic order in LiCo y Ni 1-y O 2 , a model cathode material for Li-ion batteries. In this treatment, referred to as a “spin-atom” cluster expansion, the occupant of a lattice site is identified by its spin state as well as its atomic species. Further, Effective Cluster Interaction (ECI) coefficients are derived from a large training data set (i.e., the set of atomic arrangements for which DFT calculations are performed) which is filtered by an anomaly detection algorithm to eliminate poorly converged DFT calculations. The cluster expansion incorporates Li-Ni (LN) exchange as well as intralayer Co-Ni (CN) exchange. Monte Carlo simulations based on the cluster expansion were applied to the Ni-rich part of the phase diagram. The simulations predict a miscibility gap between y = 0.05 and y = 0.65.

25 ENERGY STORAGE↗

Association of lead, cadmium and mercury with paraoxonase 1 activity and malondialdehyde in a general population in Southern Brazil

Metal exposure is associated with increased oxidative stress (OS), which is considered an underlying mechanism of metal-induced toxicity. Malondialdehyde (MDA) is a final product of lipid peroxidation, and it has been extensively used to evaluate metal-induced OS. Pro-oxidant effects produced by metals can be mitigated by paraoxonase 1 (PON1), an antioxidant enzyme known to prevent cardiovascular disease and atherosclerosis. Among other factors, the Q192R polymorphism and the exposure to heavy metals have been known to alter PON1 activity. Here, we evaluated the association of blood lead (Pb), cadmium (Cd) and mercury (Hg) levels with PON1 activity, and with MDA concentrations in a randomly selected sample of Brazilian adults aged 40 years or older, living in an urban area in Southern Brazil. A total of 889 subjects were evaluated for blood Pb and Cd levels, and 832 were tested for Hg. Geometric mean of blood Pb, Cd and Hg was 1.93 μg/dL, 0.06 μg/L and 1.40 μg/L, respectively. PON1 activity was significantly different among various genotypes: QQ (PON1=121.4 U/mL), QR (PON1=87.5 U/mL), and RR (PON1=55.2 U/mL), p<0.001. PON1 genotypes were associated only with Cd blood levels. Those with QR genotype had Cd concentrations higher (0.07 μg/L) than those with the RR genotype (0.04 μg/L) with p=0.034. However, PON1 activity was not significantly associated with metal concentrations. Cluster analysis showed that men who reported to be current smokers and drinkers with higher blood Pb and Cd levels, had significantly lower PON1 activity than non-smokers or -drinkers, and women with lower Pb and Cd levels. RR genotype carriers had lower PON1 activity than those with the QR genotype, and had higher levels of Pb and Cd compared with other genotype carriers. For blood Hg, no association with PON1 activity or genotype was noted. We found low levels of Pb, Cd and Hg in environmentally exposed Brazilian adults. Cd concentrations were increased in subjects with QR genotype. Those with RR genotype had lower PON1 activity and higher levels of Pb and Cd than other genotype carriers. The results of cluster analysis suggested that smoking status exerts a significant influence on PON1 activity. Other studies with environmentally exposed populations are required to further clarify whether low blood levels of metals influence OS biomarkers. - Highlights: • Low levels of lead, cadmium and mercury were found in a sample of Brazilian adults. • PON1 activity and MDA was not significantly associated with metals concentrations. • Blood cadmium was significantly different according with PON1 genotypes, and RR genotype carriers had lower PON1 activity and higher Pb and Cd levels. • Men, smokers, drinkers, with higher blood Pb and Cd levels had lower PON1 activity. • There was not a trend of an association with PON1 activity or genotype with blood mercury.

54 ENVIRONMENTAL SCIENCES↗

An analysis of pilot error-related aircraft accidents

A multidisciplinary team approach to pilot error-related U.S. air carrier jet aircraft accident investigation records successfully reclaimed hidden human error information not shown in statistical studies. New analytic techniques were developed and applied to the data to discover and identify multiple elements of commonality and shared characteristics within this group of accidents. Three techniques of analysis were used: Critical element analysis, which demonstrated the importance of a subjective qualitative approach to raw accident data and surfaced information heretofore unavailable. Cluster analysis, which was an exploratory research tool that will lead to increased understanding and improved organization of facts, the discovery of new meaning in large data sets, and the generation of explanatory hypotheses. Pattern recognition, by which accidents can be categorized by pattern conformity after critical element identification by cluster analysis.

Kowalsky, N. B.↗

THE SPITZER MICROLENSING PROGRAM AS A PROBE FOR GLOBULAR CLUSTER PLANETS: ANALYSIS OF OGLE-2015-BLG-0448

The microlensing event OGLE-2015-BLG-0448 was observed by Spitzer and lay within the tidal radius of the globular cluster NGC 6558. The event had moderate magnification and was intensively observed, hence it had the potential to probe the distribution of planets in globular clusters. We measure the proper motion of NGC 6558 (μ{sub cl}(N,E)=(+0.36±0.10,+1.42±0.10) mas yr{sup −1}) as well as the source and show that the lens is not a cluster member. Even though this particular event does not probe the distribution of planets in globular clusters, other potential cluster lens events can be verified using our methodology. Additionally, we find that microlens parallax measured using Optical Gravitational Lens Experiment (OGLE) photometry is consistent with the value found based on the light curve displacement between the Earth and Spitzer.

79 ASTRONOMY AND ASTROPHYSICS↗

Poisson hurdle model-based method for clustering microbiome features

Abstract Motivation High-throughput sequencing technologies have greatly facilitated microbiome research and have generated a large volume of microbiome data with the potential to answer key questions regarding microbiome assembly, structure and function. Cluster analysis aims to group features that behave similarly across treatments, and such grouping helps to highlight the functional relationships among features and may provide biological insights into microbiome networks. However, clustering microbiome data are challenging due to the sparsity and high dimensionality. Results We propose a model-based clustering method based on Poisson hurdle models for sparse microbiome count data. We describe an expectation–maximization algorithm and a modified version using simulated annealing to conduct the cluster analysis. Moreover, we provide algorithms for initialization and choosing the number of clusters. Simulation results demonstrate that our proposed methods provide better clustering results than alternative methods under a variety of settings. We also apply the proposed method to a sorghum rhizosphere microbiome dataset that results in interesting biological findings. Availability and implementation R package is freely available for download at https://cran.r-project.org/package=PHclust. Supplementary information Supplementary data are available at Bioinformatics online.

59 BASIC BIOLOGICAL SCIENCES↗

SU-F-R-33: Can CT and CBCT Be Used Simultaneously for Radiomics Analysis

Purpose: To investigate whether CBCT and CT can be used in radiomics analysis simultaneously. To establish a batch correction method for radiomics in two similar image modalities. Methods: Four sites including rectum, bladder, femoral head and lung were considered as region of interest (ROI) in this study. For each site, 10 treatment planning CT images were collected. And 10 CBCT images which came from same site of same patient were acquired at first radiotherapy fraction. 253 radiomics features, which were selected by our test-retest study at rectum cancer CT (ICC>0.8), were calculated for both CBCT and CT images in MATLAB. Simple scaling (z-score) and nonlinear correction methods were applied to the CBCT radiomics features. The Pearson Correlation Coefficient was calculated to analyze the correlation between radiomics features of CT and CBCT images before and after correction. Cluster analysis of mixed data (for each site, 5 CT and 5 CBCT data are randomly selected) was implemented to validate the feasibility to merge radiomics data from CBCT and CT. The consistency of clustering result and site grouping was verified by a chi-square test for different datasets respectively. Results: For simple scaling, 234 of the 253 features have correlation coefficient ρ>0.8 among which 154 features haveρ>0.9 . For radiomics data after nonlinear correction, 240 of the 253 features have ρ>0.8 among which 220 features have ρ>0.9. Cluster analysis of mixed data shows that data of four sites was almost precisely separated for simple scaling(p=1.29 * 10{sup −7}, χ{sup 2} test) and nonlinear correction (p=5.98 * 10{sup −7}, χ{sup 2} test), which is similar to the cluster result of CT data (p=4.52 * 10{sup −8}, χ{sup 2} test). Conclusion: Radiomics data from CBCT can be merged with those from CT by simple scaling or nonlinear correction for radiomics analysis.

60 APPLIED LIFE SCIENCES↗

BiG-SLiCE 2 v1.0.0

BiG-SLiCE was originally an open source Python-based command line bioinformatics software that offers a highly scalable clustering analysis on biosynthetic gene clusters (BGC) data. It allows a simultaneous analysis of millions of BGCs, exceeding the capability of other existing tools (around one hundred thousands). As a tradeoff, the clustering accuracy is relatively lower and sometimes fall short in corner cases and specific BGC classes such as the RiPPs (Ribosomally-translated, Post-translationally modified Peptides). In BiG-SLiCE V2 (developed in LBNL), the clustering algorithm has been significantly improved to deliver a much accurate result even for RiPPs and other previous corner case classes. Moreover, the speed of the overall pipeline has been improved by 50-100%. Finally, additional features were implemented to support downstream analyses of BiG-SLiCE results, such as customized tabular (TSV/CSV) and columnar (Parquet) outputs.

Kautsar, Satria↗

THE CLUSTERED NATURE OF STAR FORMATION. PRE-MAIN-SEQUENCE CLUSTERS IN THE STAR-FORMING REGION NGC 602/N90 IN THE SMALL MAGELLANIC CLOUD

Located at the tip of the wing of the Small Magellanic Cloud (SMC), the star-forming region NGC 602/N90 is characterized by the H II nebular ring N90 and the young cluster of pre-main-sequence (PMS) and early-type main-sequence stars NGC 602, located in the central area of the ring. We present a thorough cluster analysis of the stellar sample identified with Hubble Space Telescope/Advanced Camera for Surveys in the region. We show that apart from the central cluster low-mass PMS stars are congregated in 13 additional small, compact sub-clusters at the periphery of NGC 602, identified in terms of their higher stellar density with respect to the average background density derived from star counts. We find that the spatial distribution of the PMS stars is bimodal, with an unusually large fraction ({approx}60%) of the total population being clustered, while the remaining is diffusely distributed in the intercluster area, covering the whole central part of the region. From the corresponding color-magnitude diagrams we disentangle an age difference of {approx}2.5 Myr between NGC 602 and the compact sub-clusters, which appear younger, on the basis of comparison of the brighter PMS stars with evolutionary models, which we accurately calculated for the metal abundance of the SMC. The diffuse PMS population appears to host stars as old as those in NGC 602. Almost all detected PMS sub-clusters appear to be centrally concentrated. When the complete PMS stellar sample, including both clustered and diffused stars, is considered in our cluster analysis, it appears as a single centrally concentrated stellar agglomeration, covering the whole central area of the region. Considering also the hot massive stars of the system, we find evidence that this agglomeration is hierarchically structured. Based on our findings, we propose a scenario according to which the region NGC 602/N90 experiences an active clustered star formation for the last {approx}5 Myr. The central cluster NGC 602 was formed first and rapidly started dissolving into its immediate ambient environment, possibly ejecting also massive stars found away from its center. Star formation continued in sub-clusters of a larger stellar agglomeration, introducing an age spread of the order of 2.5 Myr among the PMS populations.

79 ASTRONOMY AND ASTROPHYSICS↗

Determining the Optimal Number of Clusters with the Clustergram

Cluster analysis aids research in many different fields, from business to biology to aerospace. It consists of using statistical techniques to group objects in large sets of data into meaningful classes. However, this process of ordering data points presents much uncertainty because it involves several steps, many of which are subject to researcher judgment as well as inconsistencies depending on the specific data type and research goals. These steps include the method used to cluster the data, the variables on which the cluster analysis will be operating, the number of resulting clusters, and parts of the interpretation process. In most cases, the number of clusters must be guessed or estimated before employing the clustering method. Many remedies have been proposed, but none is unassailable and certainly not for all data types. Thus, the aim of current research for better techniques of determining the number of clusters is generally confined to demonstrating that the new technique excels other methods in performance for several disparate data types. Our research makes use of a new cluster-number-determination technique based on the clustergram: a graph that shows how the number of objects in the cluster and the cluster mean (the ordinate) change with the number of clusters (the abscissa). We use the features of the clustergram to make the best determination of the cluster-number.

Fluegemann, Joseph K.↗

The impact of municipal budgets and land-use management on the hazardous waste production of Malaga municipalities

Economic development and the search for competitiveness have become key issues in regions' economic success. However, despite the direct relationship between economic and environmental management, few land-use plans consider the latter aspect, and city managers delegate the responsibility for environmental impacts to state legislation and private initiatives. This myopic search for competitiveness has meant that a holistic view of environmental issues is often not integrated into municipal decision-making processes. Therefore, this study's objective was to determine the relevant direct and indirect relationships of land management and budgetary procedures of municipalities with overall production levels of hazardous waste. To this end, a primary tourist region, Málaga, was examined in terms of how this waste's environmental impacts can affect the region's vital tourism sector. This research used principal component analysis, regression by ordinary least squares, cluster analysis in two stages and a means test to compare the data for the Province of Malaga's subregions. The results confirm a positive relationship between municipal expenditure and waste production and highlight the environmental benefits of land use involving environmentally non-aggressive crops. The results also reveal a negative relationship between waste production and financial assets and a direct relationship between unproductive land and the production of hazardous waste. The findings also highlight the necessity of raising awareness about the need for collaboration between different agents, especially in the development of inter-municipal strategies.

54 ENVIRONMENTAL SCIENCES↗

Quantifying and relating land-surface and subsurface variability in permafrost environments using lidar and surface geophsical datasets

The complexity of permafrost dynamics and its critical impact on climate feedbacks warrant continued development of advanced high-latitude terrestrial ecosystem characterization and monitoring approaches. In this study, we explore the value of remote sensing and surface geophysical data for characterizing land surface and subsurface properties and their linkages in an Alaskan Coastal Plain ecosystem. We base our study on data collected at the end of the 2011 growing season in the Barrow Environmental Observatory, where a nested suite of measurements were collected within a polygon-dominated region including: surface ground penetrating radar, electromagnetic, and electrical resistance tomography data; thaw depth, soil temperature and moisture content, soil texture, soil carbon and nitrogen content, and major and trace cations. Previously-collected lidar data were also available for the study. Analysis of the datasets, individually and in combination, revealed the utility of the methods for characterizing critical land-surface and subsurface properties and associated spatial zonation. Lidar analysis was performed to extract geomorphic metrics (such as slope, curvature, and directed distance of polygons), which potentially indicate drainage potential and permafrost deformation state. Cluster analysis of these lidar-obtained attributes suggested that the land surface can be grouped into three spatially coherent zones, each having a dominant geomorphic expression including: a high centered polygon zone, a low centered polygon zone and a transitional zone. Comparison of the geophysical attributes from radar, electrical resistance tomography, and electromagnetic data with point measurements suggests that the surface geophysical data can provide very high-resolution information about subsurface properties that affect ecosystem feedbacks to climate, such as thaw depth and moisture content. Cluster analysis suggested that the geophysical attributes also varied spatially in a systematic way, suggesting the presence of three laterally distinct subsurface zones. Analysis of zone-based subsurface point measurements suggests that the geophysically-defined zones have unique distributions of hydrological, thermal, and geochemical properties and that the subsurface (geophysically-based) and land-surface (lidar-based) zonation is consistent. Although the close linkage between land surface (polygonal geomorphology) and subsurface (active layer) variability revealed through our study is not surprising, to our knowledge this is the first study to document such relationships using high resolution and non-invasive approaches. This study suggests the potential of using coincident lidar and surface geophysical measurements to quantify land surface and subsurface properties (respectively) and their linkages, which are likely to play a role in terrestrial ecosystem evolution and feedbacks to climate. These findings open the way for future research focused on using combined geophysical and remote sensing datasets to estimate subsurface and land-surface properties in high resolution and over large regions as is needed for process understanding and numerical model initialization in high latitude terrestrial ecosystems.

Climate↗

Research Update: Spatially resolved mapping of electronic structure on atomic level by multivariate statistical analysis

Atomic level spatial variability of electronic structure in Fe-based superconductor FeTe0.55Se0.45 (Tc = 15 K) is explored using current-imaging tunneling-spectroscopy. Multivariate statistical analysis of the data differentiates regions of dissimilar electronic behavior that can be identified with the segregation of chalcogen atoms, as well as boundaries between terminations and near neighbor interactions. Subsequent clustering analysis allows identification of the spatial localization of these dissimilar regions. Similar statistical analysis of modeled calculated density of states of chemically inhomogeneous FeTe1 x Sex structures further confirms that the two types of chalcogens, i.e., Te and Se, can be identified by their electronic signature and differentiated by their local chemical environment. This approach allows detailed chemical discrimination of the scanning tunneling microscopy data including separation of atomic identities, proximity, and local configuration effects and can be universally applicable to chemically and electronically inhomogeneous surfaces.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗