Search NASA⌕ Search

SEARCH · Search NASA

Results for “Spectral Cluster”

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

Measurement of the spectral signature of small carbon clusters at near and far infrared wavelengths

A significant percentage of the carbon inventory of the circumstellar and interstellar media may be in the form of large refractory molecules (or small grains) referred to as carbon clusters. At the small end, uneven numbers of carbon atoms seem to be preferred, whereas above 12 atoms, clusters containing an even number of carbon atoms appear to be preferred in laboratory chemistry. In the lab, the cluster C-60 appears to be a particularly stable form and has been nicknamed Bucky Balls because of its resemblance to a soccer ball and to geodesic domes designed by Buckminster Fuller. In order to investigate the prevalence of these clusters, and their relationship to the polycyclic aromatic hydrocarbons (PAHs) that have become the newest focus of IR astronomy, it is necessary to determine the spectroscopic characteristics of these clusters at near and far infrared wavelengths. Described here is the construction of a near to far IR laser magnetic resonance spectrometer that has been built at the University of California Berkeley in order to detect and characterize these spectra. The equipment produces carbon clusters by laser evaporation of a graphitic target. The clusters are then cooled in a supersonic expansion beam in order to simulate conditions in the interstellar medium (ISM). The expansion beam feeds into the spectrometer chamber and permits concentrations of clusters sufficiently high as to permit ultra-high resolution spectroscopy at near and far IR wavelengths. The first successful demonstration of this apparatus occurred last year when the laboratory studies permitted the observational detection of C-5 in the stellar outflow surrounding IRC+10216 in the near-IR. Current efforts focus on reducing the temperature of the supersonic expansion beam that transport the C clusters evaporated from a graphite target into the spectrometer down to temperatures as low as 1 K.

Tarter, J.↗

BOREAS AFM-12 1-km AVHRR Seasonal Land Cover Classification

The Boreal Ecosystem-Atmosphere Study (BOREAS) Airborne Fluxes and Meteorology (AFM)-12 team's efforts focused on regional scale Surface Vegetation and Atmosphere (SVAT) modeling to improve parameterization of the heterogeneous BOREAS landscape for use in larger scale Global Circulation Models (GCMs). This regional land cover data set was developed as part of a multitemporal one-kilometer Advanced Very High Resolution Radiometer (AVHRR) land cover analysis approach that was used as the basis for regional land cover mapping, fire disturbance-regeneration, and multiresolution land cover scaling studies in the boreal forest ecosystem of central Canada. This land cover classification was derived by using regional field observations from ground and low-level aircraft transits to analyze spectral-temporal clusters that were derived from an unsupervised cluster analysis of monthly Normalized Difference Vegetation Index (NDVI) image composites (April-September 1992). This regional data set was developed for use by BOREAS investigators, especially those involved in simulation modeling, remote sensing algorithm development, and aircraft flux studies. Based on regional field data verification, this multitemporal one-kilometer AVHRR land cover mapping approach was effective in characterizing the biome-level land cover structure, embedded spatially heterogeneous landscape patterns, and other types of key land cover information of interest to BOREAS modelers.The land cover mosaics in this classification include: (1) wet conifer mosaic (low, medium, and high tree stand density), (2) mixed coniferous-deciduous forest (80% coniferous, codominant, and 80% deciduous), (3) recent visible bum, vegetation regeneration, or rock outcrops-bare ground-sparsely vegetated slow regeneration bum (four classes), (4) open water and grassland marshes, and (5) general agricultural land use/ grasslands (three classes). This land cover mapping approach did not detect small subpixel-scale landscape features such as fens, bogs, and small water bodies. Field observations and comparisons with Landsat Thematic Mapper (TM) suggest a minimum effective resolution of these land cover classes in the range of three to four kilometers, in part, because of the daily to monthly compositing process. In general, potential accuracy limitations are mitigated by the use of conservative parameterization rules such as aggregation of predominant land cover classes within minimum horizontal grid cell sizes of ten kilometers. The AFM-12 one-kilometer AVHRR seasonal land cover classification data are available from the Earth Observing System Data and Information System (EOSDIS) Oak Ridge National Laboratory (ORNL) Distributed Active Archive Center (DAAC). The data files are available on a CD-ROM (see document number 20010000884).

Steyaert, Lou↗

The Atacama Cosmology Telescope: Cosmological Parameters from the 2008 Power Spectrum

We present cosmological parameters derived from the angular power spectrum of the cosmic microwave background (CMB) radiation observed at 148 GHz and 218 GHz over 296 deg(exp 2) with the Atacama Cosmology Telescope (ACT) during its 2008 season. ACT measures fluctuations at scales 500 < l < 10,000. We fit a model for the lensed CMB, Sunyaev-Zel'dovich (SZ), and foreground contribution to the 148 GHz and 218 GHz power spectra, including thermal and kinetic SZ, Poisson power from radio and infrared point sources, and clustered power from infrared point sources. At l = 3000, about half the power at 148 GHz comes from primary CMB after masking bright radio sources. The power from thermal and kinetic SZ is estimated to be Beta(sub 3000) is identical to 6.8 +/- 2.9 mu K (exp 2), where Beta (sub l) is identical to l(l + 1) C(sub l)/2pi. The IR Poisson power at 148 GHz is Bewta(sub 3000) 7.8 +/- 0.7 muK(exp 2) (C(sub l) = 5.5 +/- 0.5 nK(exp 2)), and a clustered IR component is required with Beta (sub 3000) = 4.6 +/- 0.9 muK(exp 2), assuming an analytic model for its power spectrum shape. At 218 GHz only about 15% of the power, approximately 27 mu K(exp 2), is CMB anisotropy at l = 3000. The remaining 85% is attributed to IR sources (approximately 50% Poisson and 35% clustered), with spectral index alpha = 3.69 +/- 0.14 for flux scaling as S(nu) varies as nu(sup alpha). We estimate primary cosmological parameters from the less contaminated 148 GHz spectrum, marginalizing over SZ and source power. The ACDM cosmological model is a good fit to the data (chi square/dof = 29/46), and ACDM parameters estimated from ACT+Wilkinson Microwave Anisotropy Probe (WMAP) are consistent with the seven-year WMAP limits, with scale invariant n(sub s) = 1 excluded at 99.7% confidence level (CL) (3 sigma). A model with no CMB lensing is disfavored at 2.8 sigma. By measuring the third to seventh acoustic peaks, and probing the Silk damping regime, the ACT data improve limits on cosmological parameters that affect the small-scale CMB power. The ACT data combined with WMAP give a 6 sigma detection of primordial helium, with Y(sub p) = 0.313 +/- 0.044, and a 4 sigma detection of relativistic species, assumed to be neutrinos, with N(sub eff) = 5.3 +/- 1.3 (4.6 +/- 0.8 with BAO+H(sub 0) data). From the CMB alone the running of the spectral index is constrained to be d(sub s) / d ln k = -0,034 +/- 0,018, the limit on the tensor-to-scalar ratio is r < 0,25 (95% CL), and the possible contribution of Nambu cosmic strings to the power spectrum is constrained to string tension G(sub mu) < 1.6 x 10(exp -7) (95% CL),

Dunkley, J.↗

XRISM Constraints on Unidentified X-Ray Emission Lines, Including the 3.5 keV Line, in the Stacked Spectrum of 10 Galaxy Clusters

We stack 3.75 Ms of early XRISM Resolve observations of 10 galaxy clusters to search for unidentified spectral lines in the E = 2.5–15 keV band (rest frame), including the E = 3.5 keV line reported in earlier low spectral resolution studies of cluster samples. Such an emission line may originate from the decay of the sterile neutrino, a warm dark matter (DM) candidate. No unidentified lines are detected in our stacked cluster spectrum, with the 3σ upper limit on the m s ∼ 7.1 keV DM particle decay rate (which corresponds to an E = 3.55 keV emission line) of Γ ∼ 1.0 × 10 −27 s −1 . This upper limit is 3–4 times lower than the one derived by Hitomi Collaboration from the Perseus observation but still 5 times higher than the XMM-Newton detection reported by E. Bulbul et al. in the stacked cluster sample. XRISM Resolve, with its high spectral resolution but small field of view, may reach the sensitivity needed to test the XMM-Newton cluster sample detection by combining several years worth of future cluster observations.

Astronomy and AstroPhysics↗

Exciton energy transfer reveals spectral signatures of excited states in clusters

Electronic excitation and concomitant energy transfer leading to Penning ionization in argon–acetylene clusters generated in a supersonic expansion are investigated with synchrotron-based photoionization mass spectrometry and electronic structure calculations. Spectral features in the photoionization efficiency of the mixed argon–acetylene clusters reveal a blue shift from the 2 P 1/2 and 2 P 3/2 excited states of atomic argon. Analysis of this feature suggests that excited states of argon clusters transfer energy to acetylene, resulting in its ionization and successive evaporation of argon. Theoretically calculated Ar n (n = 2–6) cluster spectra are in excellent agreement with experimental observations, and provide insight into the structure and ionization dynamics of the clusters. A comparison between argon–acetylene and argon–water clusters reveals that argon solvates water better, allowing for higher-order excitons and Rydberg states to be populated. Furthermore, these results are explained by theoretical calculations of respective binding energies and structures.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

X-ray spectra of 12 distant clusters of galaxies

An X-ray spectral analysis of 12 distant clusters of galaxies observed on-axis with the Einstein Observatory imaging proportional counter is presented. Statistically, X-ray spectral data of the highest red shift clusters in the sample are inconsistent with an optically thin plasma emission model plus absorptions only in the Milky Way. Inclusions of excess absorptions in the clusters significantly improve the spectral fits to the data, indicating the possible presence of large amounts of X-ray absorbing cool gas in some of the distant clusters. By using X-ray luminosities determined from observed fluxes only in the 0.8-3.5 keV band and the temperature-to-luminosity correlation for nearby clusters, the estimated temperatures of the hot intracluster medium in these distant clusters constrain the absorption columns in the clusters. The data also show possible positive temperature evolution of the hot ICM with time which, if present, would further increase the absorptions required in fitting the spectra.

Wang, Qingde↗

Helium-Like Iron Line Temperature Diagnostics in Clusters of Galaxies

The emission complex around 6.7 keV, arising from He-like Fe lines and their dielectronic satellites, has been observed at low resolution in a number of clusters of galaxies. With sufficient spectral resolution as in, e.g., solar flare observations, the ratio, G = (x + y + z)/w, of the intercombination (x and y) and forbidden (z) to the resonance (w) lines arising from the n = 2 level of the He-like ion, is a sensitive temperature diagnostic. We examine this role for G in the context of spatially and spectrally resolved observations of clusters of galaxies. The temperature behavior of this ratio is strongly dependent on the spectral resolution. We introduce an alternative definition for G that includes the contribution of satellite lines and improves the temperature fidelity of this diagnostic. We find that deprojection of the observed value of G along a line of sight through the cluster can yield accurate temperature profiles for temperatures in the range 10(exp 7)-10(exp 8) K that do not suffer from the cluster model uncertainties inherent in deconvolution of broad-band X-ray surface brightness profiles.

Swartz, Douglas A.↗

X-ray emission from clusters of galaxies

Some X-ray spectral observations of approximately 30 clusters of galaxies from HEAO-1 are summarized. There exists strong correlations between X-ray luminosity, L(x), and temperature kT in the form L(x)alphaT to the 2.3 power. This result combined with the L(x) central galaxy density relation and the virial theorem indicates that the core dadius of the gas should be roughly independent of L(x) or KT and that more luminous clusters have a greater fraction of their virial mass in gas. The poor correlation of KT and optical velocity dispersion seems to indicate that clusters have a variety of equations of state. There is poor agreement between X-ray imaging observations and optical and X-ray spectral measures of the polytropic index. Most clusters show Fe emission lines with a strong indication that they all have roughly 1/2 solar abundance. The evidence for cooling in the cores of several clusters is discussed based on spectral observations with the Einstein solid state spectrometer.

Mushotzky, R. F.↗

X-ray emission from clusters of galaxies

Some X-ray spectral observations of approximately 30 clusters of galaxies from HEAO-1 are summarized. There exists strong correlation betwen X-ray luminosity, L(x), and temperature kT in the form L(x)alphaT to the 2.3 power. This result combined with the L(x) central galaxy density relation and the virial theorem indicates that the core dadius of the gas should be roughly independent of L(x) or Kt and that more luminous clusters have a greater fraction of their virial mass in gas. The poor correlation of KT and optical velocity dispersion seems to indicate that clusters have a variety of equations of state. There is poor agreement between X-ray imaging observations and optical and X-ray spectral measures of the polytropic index. Most clusters show Fe emission lines with a strong indication that they all have roughly 1/2 solar abundance. The evidence for cooling in the cores several clusters is discussed based on spectral observations with the Einstein solid state spectrometer.

Mushotzky, R. F.↗

Model Order Reduction Algorithm for Estimating the Absorption Spectrum

The ab initio description of the spectral interior of the absorption spectrum poses both a theoretical and computational challenge for modern electronic structure theory. Due to the often spectrally dense character of this domain in the quantum propagator’s eigenspectrum for medium-to-large sized systems, traditional approaches based on the partial diagonalization of the propagator often encounter oscillatory and stagnating convergence. Electronic structure methods which solve the molecular response problem through the solution of spectrally shifted linear systems, such as the complex polarization propagator, offer an alternative approach which is agnostic to the underlying spectral density or domain location. This generality comes at a seemingly high computational cost associated with solving a large linear system for each spectral shift in some discretization of the spectral domain of interest. In this work, we present a novel, adaptive solution to this high computational overhead based on model order reduction techniques via interpolation. Model order reduction reduces the computational complexity of mathematical models and is ubiquitous in the simulation of dynamical systems and control theory. The efficiency and effectiveness of the proposed algorithm in the ab initio prediction of X-ray absorption spectra is demonstrated using a test set of challenging water clusters which are spectrally dense in the neighborhood of the oxygen K-edge. On the basis of a single, user defined tolerance we automatically determine the order of the reduced models and approximate the absorption spectrum up to the given tolerance. We also illustrate that, for the systems studied, the automatically determined model order increases logarithmically with the problem dimension, compared to a linear increase of the number of eigenvalues within the energy window. Furthermore, we observed that the computational cost of the proposed algorithm only scales quadratically with respect to the problem dimension.

IDREAM, K-Edge, theoretical, computational, electr↗

Planck intermediate results: XLIII. Spectral energy distribution of dust in clusters of galaxies

Although infrared (IR) overall dust emission from clusters of galaxies has been statistically detected using data from the Infrared Astronomical Satellite (IRAS), it has not been possible to sample the spectral energy distribution (SED) of this emission over its peak, and thus to break the degeneracy between dust temperature and mass. By complementing the IRAS spectral coverage with Planck satellite data from 100 to 857 GHz, we provide in this paper new constraints on the IR spectrum of thermal dust emission in clusters of galaxies. We achieve this by using a stacking approach for a sample of several hundred objects from the Planck cluster sample. This procedure averages out fluctuations from the IR sky, allowing us to reach a significant detection of the faint cluster contribution. We also use the large frequency range probed by Planck, together with component-separation techniques, to remove the contamination from both cosmic microwave background anisotropies and the thermal Sunyaev-Zeldovich effect (tSZ) signal, which dominate at ν ≤ 353 GHz. By excluding dominant spurious signals or systematic effects, averaged detections are reported at frequencies 353 GHz ≤ ν ≤ 5000 GHz. We confirm the presence of dust in clusters of galaxies at low and intermediate redshifts, yielding an SED with a shape similar to that of the Milky Way. Planck’s resolution does not allow us to investigate the detailed spatial distribution of this emission (e.g. whether it comes from intergalactic dust or simply the dust content of the cluster galaxies), but the radial distribution of the emission appears to follow that of the stacked SZ signal, and thus the extent of the clusters. Finally, the recovered SED allows us to constrain the dust mass responsible for the signal and its temperature.

79 ASTRONOMY AND ASTROPHYSICS↗

Molecular-like Ag clusters sensitized near-infrared down-conversion luminescence in oxyfluoride glasses for broadband spectral modification

Molecular-like Ag clusters sized at 1–4 nm have been stabilized in Pb/Cd-free oxyfluoride glasses, showing broadband excitation/emission characteristics and unique wavelength-dependent luminescent performance with a maximal quantum yield of 26.9%. It was experimentally demonstrated that an energy transfer route of Ag clusters → Tb{sup 3+} → Yb{sup 3+} occurs in Ag{sup +}/Tb{sup 3+}/Yb{sup 3+} tri-doped sample, wherein Ag clusters act as sensitizers for near-infrared down-conversion spectral modification. Hopefully, the proposed strategy that noble metal clusters being applied for harvesting solar radiation may potentially solve the sticky problems of the narrow excitation bandwidth and the low excitation efficiency in rare earth ions doped down-conversion materials.

36 MATERIALS SCIENCE↗

Probing the early stages of salt nucleation—Experimental and theoretical investigations of sodium/potassium thiocyanate cluster anions

Due to the fast solvent evaporation in electrospray ionization (ESI), the concentration of initially dilute electrolyte solutions rapidly increases to afford the formation of supersaturated droplets and generation of various pristine anhydrous salt clusters in the gas phase. The size, composition, and charge distributions of these clusters, in principle, witness the nucleation evolution in solutions. Herein, we report a microscopic study on the initial stage of nucleation and crystallization of sodium/potassium thiocyanate salt solutions simulated in the ESI process. Singly charged M{sub x}(SCN){sub x+1}{sup −}, doubly charged M{sub y}(SCN){sub y+2}{sup 2−} (M = Na, K), and triply charged K{sub z}(SCN){sub z+3}{sup 3−} anion clusters (x, y, and z stand for the number of alkali atoms in the singly, doubly, and triply charged clusters, respectively) were produced via electrospray of the corresponding salt solutions and were characterized by negative ion photoelectron spectroscopy (NIPES). The vertical detachment energies (VDEs) of these sodium/potassium thiocyanate cluster anions were obtained, and theoretical calculations were carried out for the sodium thiocyanate clusters in assisting spectral identification. The measured VDEs of singly charged anions M{sub x}(SCN){sub x+1}{sup −} (M = Na and K) demonstrate that they are superhalogen anions. The existence of doubly charged anions M{sub y}(SCN){sub y+2}{sup 2−} (y = 2x, x ≥ 4 and 3 for M = Na and K, respectively) and triply charged anions K{sub z}(SCN){sub z+3}{sup 3−} (z = 3x, x ≥ 6) was initially discovered from the photoelectron spectra for those singly charged anions of M{sub x}(SCN){sub x+1}{sup −} with the same mass-to-charge ratio (m/z), and later independently confirmed by the observation of their distinct mass spectral distributions and by taking their NIPE spectra for those pure multiply charged anions with their m/z different from the singly charged species. For large clusters, multiply charged clusters were found to become preferred, but at higher temperatures, those multiply charged clusters were suppressed. The series of anion clusters investigated here range from molecular-like M{sub 1}(SCN){sub 2}{sup −} to nano-sized K{sub 22}(SCN){sub 25}{sup 3−}, providing a vivid molecular-level growth pattern reflecting the initial salt nucleation process.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

AN ENHANCED SPECTROSCOPIC CENSUS OF THE ORION NEBULA CLUSTER

We report new spectral types or spectral classification constraints for over 600 stars in the Orion Nebula Cluster (ONC) based on medium resolution (R ≈ 1500-2000) red optical spectra acquired using the Palomar 200'' and Kitt Peak 3.5 m telescopes. Spectral types were initially estimated for F, G, and early K stars from atomic line indices while for late K and M stars, which constitute the majority of our sample, indices involving TiO and VO bands were used. To ensure proper classification, particularly for reddened, veiled, or nebula-contaminated stars, all spectra were then visually examined for type verification or refinement. We provide an updated spectral type table that supersedes previous work, increasing the percentage of optically visible ONC stars with spectral type information from 68% to 90%. However, for many objects, repeated observations have failed to yield spectral types primarily due to the challenges of adequate sky subtraction against a bright and spatially variable nebular background. The scatter between our new and our previously determined spectral types is approximately two spectral sub-classes. We also compare our grating spectroscopy results with classification based on narrow-band TiO filter photometry, finding similar scatter. While the challenges of working in the ONC may explain much of the spread, we highlight several stars showing significant and unexplained bona fide spectral variations in observations taken several years apart; these and similar cases could be due to a combination of accretion and extinction changes. Finally, nearly 20% of ONC stars exhibit obvious Ca II triplet emission indicative of strong accretion.

79 ASTRONOMY AND ASTROPHYSICS↗

Deep Very Large Array Observations of the Merging Cluster CIZA J2242.8+5301: Continuum and Spectral Imaging

Despite the progress that has been made in understanding radio relics, there are still open questions regarding the underlying particle acceleration mechanisms. In this paper, we present deep 1–4 GHz Very Large Array (VLA) observations of CIZA J2242.8+5301 ( z = 0.1921), a double radio relic cluster characterized by small projection on the plane of the sky. Our VLA observations reveal, for the first time, the complex morphology of the diffuse sources and the filamentary structure of the northern relic. Additionally, we discover new, faint, diffuse radio emission extending north of the main northern relic. Our Mach number estimates for the northern and southern relics, based on the radio spectral index map obtained using the VLA observations and existing LOw Frequency ARray (LOFAR) and Giant Metrewave Radio Telescope data, are consistent with previous radio and X-ray studies ( and ). However, color–color diagrams and models suggest a flatter injection spectral index than the one obtained from the spectral index map, indicating that projection effects might be not entirely negligible. The southern relic consists of five "arms." Embedded in it, we find a tailed radio galaxy that seems to be connected to the relic. A spectral index flattening, where the radio tail connects to the relic, is also measured. We propose that the southern relic may trace AGN fossil electrons that are reaccelerated by a shock, with an estimated strength of . High-resolution mapping of other tailed radio galaxies also supports a scenario where AGN fossil electrons are revived by the merger event and could be related to the formation of some diffuse cluster radio emission.

79 ASTRONOMY AND ASTROPHYSICS↗

Placing the Spotted T Tauri Star LkCa 4 on an HR Diagram

Ages and masses of young stars are often estimated by comparing their luminosities and effective temperatures to pre-main-sequence stellar evolution tracks, but magnetic fields and starspots complicate both the observations and evolution. To understand their influence, we study the heavily spotted weak-lined T-Tauri star LkCa 4 by searching for spectral signatures of radiation originating from the starspot or starspot groups. We introduce a new methodology for constraining both the starspot filling factor and the spot temperature by fitting two-temperature stellar atmosphere models constructed from Phoenix synthetic spectra to a high-resolution near-IR IGRINS spectrum. Clearly discernable spectral features arise from both a hot photospheric component T{sub hot} ∼ 4100 K and a cool component T{sub cool} ∼ 2700–3000 K, which covers ∼80% of the visible surface. This mix of hot and cool emission is supported by analyses of the spectral energy distribution, rotational modulation of colors and of TiO band strengths, and features in low-resolution optical/near-IR spectroscopy. Although the revised effective temperature and luminosity make LkCa 4 appear to be much younger and of much lower mass than previous estimates from unspotted stellar evolution models, appropriate estimates will require the production and adoption of spotted evolutionary models. Biases from starspots likely afflict most fully convective young stars and contribute to uncertainties in ages and age spreads of open clusters. In some spectral regions, starspots act as a featureless “veiling” continuum owing to high rotational broadening and heavy line blanketing in cool star spectra. Some evidence is also found for an anticorrelation between the velocities of the warm and cool components.

79 ASTRONOMY AND ASTROPHYSICS↗

X-ray spectral diagnostics of laser harmonic generation in the interaction of relativistic femtosecond laser pulses with clusters

It is shown that the production of X-ray emission spectra in the interaction of high-intensity laser radiation with cluster targets may be affected by the bichromatic oscillating electric field arising from the generation of the second harmonic of laser radiation. A technique is proposed for diagnosing harmonic generation in laser – cluster interactions using the spectral line profiles of multiply charged helium ions. The efficiency of second harmonic generation at a laser intensity of 3 × 10{sup 18} W cm{sup -2} is shown to amount to about 2%. (extreme laser radiation: physics and fundamental applications)

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Significant DBSCAN+: Statistically Robust Density-based Clustering

Cluster detection is important and widely used in a variety of applications, including public health, public safety, transportation, and so on. Given a collection of data points, we aim to detect density-connected spatial clusters with varying geometric shapes and densities, under the constraint that the clusters are statistically significant. The problem is challenging, because many societal applications and domain science studies have low tolerance for spurious results, and clusters may have arbitrary shapes and varying densities. As a classical topic in data mining and learning, a myriad of techniques have been developed to detect clusters with both varying shapes and densities (e.g., density-based, hierarchical, spectral, or deep clustering methods). However, the vast majority of these techniques do not consider statistical rigor and are susceptible to detecting spurious clusters formed as a result of natural randomness. On the other hand, scan statistic approaches explicitly control the rate of spurious results, but they typically assume a single “hotspot” of over-density and many rely on further assumptions such as a tessellated input space. To unite the strengths of both lines of work, we propose a statistically robust formulation of a multi-scale DBSCAN, namely Significant DBSCAN+, to identify significant clusters that are density connected. As we will show, incorporation of statistical rigor is a powerful mechanism that allows the new Significant DBSCAN+ to outperform state-of-the-art clustering techniques in various scenarios. We also propose computational enhancements to speed-up the proposed approach. Experiment results show that Significant DBSCAN+ can simultaneously improve the success rate of true cluster detection (e.g., 10–20% increases in absolute F1 scores) and substantially reduce the rate of spurious results (e.g., from thousands/hundreds of spurious detections to none or just a few across 100 datasets), and the acceleration methods can improve the efficiency for both clustered and non-clustered data.

Computer Science↗