Search NASA⌕ Search

SEARCH · Search NASA

Results for “sampling algorithms”

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 1,261 records · Page 70

An optimization of the FPGA trigger based on the artificial neural network for a detection of neutrino-origin showers

Observations of ultra-high energy neutrinos became a priority in experimental astro-particle physics. Up to now, the Pierre Auger Observatory did not find any candidate on a neutrino event. This imposes competitive limits to the diffuse flux of ultra-high energy neutrinos in the EeV range and above. A very low rate of events potentially generated by neutrinos is a significant challenge for a detection technique and requires both sophisticated algorithms and high-resolution hardware. A trigger based on a artificial neural network was implemented into the Cyclone{sup R} V E FPGA 5CEFA9F31I7. The prototype Front-End boards for Auger-Beyond-2015 with Cyclone{sup R} V E can test the neural network algorithm in real pampas conditions in 2015. Showers for muon and tau neutrino initiating particles on various altitudes, angles and energies were simulated in CORSICA and Offline platforms giving pattern of ADC traces in Auger water Cherenkov detectors. The 3-layer 12-10-1 neural network was taught in MATLAB by simulated ADC traces according the Levenberg-Marquardt algorithm. Results show that a probability of a ADC traces generation is very low due to a small neutrino cross-section. Nevertheless, ADC traces, if occur, for 1-10 EeV showers are relatively short and can be analyzed by 16-point input algorithm. For 100 EeV range traces are much longer, but with significantly higher amplitudes, which can be detected by standard threshold algorithms. We optimized the coefficients from MATLAB to get a maximal range of potentially registered events and for fixed-point FPGA processing to minimize calculation errors. Currently used Front-End boards based on no-more produced ACEXR PLDs and obsolete Cyclone{sup R} FPGAs allow an implementation of relatively simple threshold algorithms for triggers. New sophisticated trigger implemented in Cyclone{sup R} V E FPGAs with large amount of DSP blocks, embedded memory running with 120 - 160 MHz sampling may support to discover neutrino events in the Pierre Auger Observatory. (authors)

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

AUTOMATIC DETERMINATION OF STELLAR PARAMETERS VIA ASTEROSEISMOLOGY OF STOCHASTICALLY OSCILLATING STARS: COMPARISON WITH DIRECT MEASUREMENTS

Space-based projects are providing a wealth of high-quality asteroseismic data, including frequencies for a large number of stars showing solar-like oscillations. These data open the prospect for precise determinations of key stellar parameters, of particular value to the study of extra-solar planetary systems. Given the quantity of the available and expected data, it is important to develop efficient and reliable techniques for analyzing them, including the determination of stellar parameters from the observed frequencies. Here we present the SEEK package developed for the analysis of asteroseismic data from the Kepler mission. A central goal of the package is to obtain a fast and automatic determination of the stellar radius and other parameters in a form that is statistically well defined. The algorithms are tested by comparing the results of the analysis with independent measurements of stellar radius and mass for a sample of well-observed stars. We conclude that the SEEK package fixes stellar parameters with accuracy and precision.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Rare Events via Cross-Entropy Population Monte Carlo

Rare events are events that happen with very low frequency. Estimating rare event probabilities using Monte Carlo techniques is computationally expensive, often to the point of intractability, and special methods are required. Importance sampling (IS) is a well known technique that uses a proposal distribution in place of a target distribution to lower the variance of the estimator. Key to the success of IS methods is the choice of a proposal distribution, or the parameters governing the distribution. Adaptive importance sampling improves the parameters of a family or population of proposal distributions iteratively through trials. We present a novel cross-entropy population Monte Carlo algorithm, which adapts the parameters of proposals through the cross-entropy method. The proposed method stands apart from previous work in that we are not optimizing a mixture distribution. Instead, we leverage deterministic mixture weights and optimize the distributions individually through a reinterpretation of the typical derivation of the cross-entropy method. Demonstrations on rare event examples show that the algorithm can outperform existing resampling based population Monte Carlo methods, especially for higher-dimensional problems. Finally, we also demonstrate efficacy on a conjunction analysis problem.

97 MATHEMATICS AND COMPUTING↗

Simulation and Optimization of Large Scale Subsurface Environmental Impacts; Investigations, Remedial Design and Long Term Monitoring

The global impact to human health and the environment from large scale chemical / radionuclide releases is well documented. Examples are the wide spread release of radionuclides from the Chernobyl nuclear reactors, the mobilization of arsenic in Bangladesh, the formation of Environmental Protection Agencies in the United States, Canada and Europe, and the like. The fiscal costs of addressing and remediating these issues on a global scale are astronomical, but then so are the fiscal and human health costs of ignoring them. An integrated methodology for optimizing the response(s) to these issues is needed. This work addresses development of optimal policy design for large scale, complex, environmental issues. It discusses the development, capabilities, and application of a hybrid system of algorithms that optimizes the environmental response. It is important to note that 'optimization' does not singularly refer to cost minimization, but to the effective and efficient balance of cost, performance, risk, management, and societal priorities along with uncertainty analysis. This tool integrates all of these elements into a single decision framework. It provides a consistent approach to designing optimal solutions that are tractable, traceable, and defensible. The system is modular and scalable. It can be applied either as individual components or in total. By developing the approach in a complex systems framework, a solution methodology represents a significant improvement over the non-optimal 'trial and error' approach to environmental response(s). Subsurface environmental processes are represented by linear and non-linear, elliptic and parabolic equations. The state equations solved using numerical methods include multi-phase flow (water, soil gas, NAPL), and multicomponent transport (radionuclides, heavy metals, volatile organics, explosives, etc.). Genetic programming is used to generate the simulators either when simulation models do not exist, or to extend the accuracy of them. The uncertainty and sparse nature of information in earth science simulations necessitate stochastic representations. For discussion purposes, the solution to these site-wide challenges is divided into three sub-components; plume finding, long term monitoring, and site-wide remediation. Plume finding is the optimal estimation of the plume fringe(s) at a specified time. It is optimized by fusing geo-stochastic flow and transport simulations with the information content of data using a Kalman filter. The result is an optimal monitoring sensor network; the decision variable is location(s) of sensor in three dimensions. Long term monitoring extends this approach concept, and integrates the spatial-time correlations to optimize the decision variables of where to sample and when to sample over the project life cycle. Optimization of location and timing of samples to meet the desired accuracy of temporal plume movement is accomplished using enumeration or genetic algorithms. The remediation optimization solves the multi-component, multiphase system of equations and incorporates constraints on life-cycle costs, maximum annual costs, maximum allowable annual discharge (for assessing the monitored natural attenuation solution) and constraints on where remedial system component(s) can be located, including management overrides to force certain solutions to be chosen are incorporated for solution design. It uses a suite of optimization techniques, including the outer approximation method, Lipchitz global optimization, genetic algorithms, and the like. The automated optimal remedial design algorithm requires a stable simulator be available for the simulated process. This is commonly the case for all above specifications sans true three-dimensional multiphase flow. Much work is currently being conducted in the industry to develop stable 3D, three-phase simulators. If needed, an interim heuristic algorithm is available to get close to optimal for these conditions. This system process provides the full capability to optimize multi-source, multiphase, and multicomponent sites. The results of applying just components of these algorithms have produced predicted savings of as much as $90,000,000(US), when compared to alternative solutions. Investment in a pilot program to test the model saved 100% of the $20,000,000 predicted for the smaller test implementation. This was done without loss of effectiveness, and received an award from the Vice President - and now Nobel peace prize winner - Al Gore of the United States. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Measurement of inclusive jet and dijet production in pp collisions at root s=7 TeV using the ATLAS detector

Inclusive jet and dijet cross sections have been measured in proton-proton collisions at a center-of-mass energy of 7 TeV using the ATLAS detector at the Large Hadron Collider. The cross sections were measured using jets clustered with the anti-k{sub t} algorithm with parameters R = 0.4 and R = 0.6. These measurements are based on the 2010 data sample, consisting of a total integrated luminosity of 37 pb{sup -1}. Inclusive jet double-differential cross sections are presented as a function of jet transverse momentum, in bins of jet rapidity. Dijet double-differential cross sections are studied as a function of the dijet invariant mass, in bins of half the rapidity separation of the two leading jets. The measurements are performed in the jet rapidity range |y| < 4.4, covering jet transverse momenta from 20 GeV to 1.5 TeV and dijet invariant masses from 70 GeV to 5 TeV. The data are compared to expectations based on next-to-leading-order QCD calculations corrected for nonperturbative effects, as well as to next-to-leading-order Monte Carlo predictions. In addition to a test of the theory in a new kinematic regime, the data also provide sensitivity to parton distribution functions in a region where they are currently not well-constrained.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Identification of heavy-flavour jets with the CMS detector in pp collisions at 13 TeV

Many measurements and searches for physics beyond the standard model at the LHC rely on the efficient identification of heavy-flavour jets, i.e. jets originating from bottom or charm quarks. In this paper, the discriminating variables and the algorithms used for heavy-flavour jet identification during the first years of operation of the CMS experiment in proton-proton collisions at a centre-of-mass energy of 13 TeV, are presented. Heavy-flavour jet identification algorithms have been improved compared to those used previously at centre-of-mass energies of 7 and 8 TeV. For jets with transverse momenta in the range expected in simulated events, these new developments result in an efficiency of 68% for the correct identification of a b jet for a probability of 1% of misidentifying a light-flavour jet. The improvement in relative efficiency at this misidentification probability is about 15%, compared to previous CMS algorithms. In addition, for the first time algorithms have been developed to identify jets containing two b hadrons in Lorentz-boosted event topologies, as well as to tag c jets. The large data sample recorded in 2016 at a centre-of-mass energy of 13 TeV has also allowed the development of new methods to measure the efficiency and misidentification probability of heavy-flavour jet identification algorithms. The b jet identification efficiency is measured with a precision of a few per cent at moderate jet transverse momenta (between 30 and 300 GeV) and about 5% at the highest jet transverse momenta (between 500 and 1000 GeV).

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Inversion of Many-Beam Bragg Intensities for Phasing by Iterated Projections: Removal of Multiple Scattering Artifacts from Diffraction Data

An iterated projection algorithm (N-Phaser) is developed that reconstructs a scattering potential from N-beam multiple Bragg scattered intensities. The method may be used to eliminate multiple scattering artifacts from electron diffraction data, solving the phase problem and increasing the thicknesses of samples used in materials science, solid-state chemistry, and small molecule crystallography. For high-energy transmission electron diffraction, we show that the algorithm recovers accurate complex structure factors from a wide range of thicknesses, orientations, and relativistic beam energies, and does not require known thickness or atomic-resolution data if sufficient multiple scattering occurs. Furthermore, extensions to Cryo-electron microscopy and Micro-electron diffraction are suggested.

36 MATERIALS SCIENCE↗

An architecture for real time data acquisition and online signal processing for high throughput tandem mass spectrometry

Independent, greedy collection of data events using simple heuristics results in massive over-sampling of the prominent data features in large-scale studies over what should be achievable through “intelligent,” online acquisition of such data. As a result, data generated are more aptly described as a collection of a large number of small experiments rather than a true large-scale experiment. Nevertheless, achieving “intelligent,” online control requires tight interplay between state-of-the-art, data-intensive computing infrastructure developments and analytical algorithms. In this paper, we propose a Software Architecture for Mass spectrometry-based Proteomics coupled with Liquid chromatography Experiments (SAMPLE) to develop an “intelligent” online control and analysis system to significantly enhance the information content from each sensor (in this case, a mass spectrometer). Using online analysis of data events as they are collected and decision theory to optimize the collection of events during an experiment, we aim to maximize the information content generated during an experiment by the use of pre-existing knowledge to optimize the dynamic collection of events.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Validation of a Custom-made Software for DQE Assessment in Mammography Digital Detectors

This works presents the validation of a custom-made software, designed and developed in Matlab, intended for routine evaluation of detective quantum efficiency DQE, according to algorithms described in the IEC 62220-1-2 standard. DQE, normalized noise power spectrum NNPS and pre-sampling modulation transfer function MTF were calculated from RAW images from a GE Senographe DS (FineView disabled) and a Siemens Novation system. Calculated MTF is in close agreement with results obtained with alternative codes: MTF lowbar tool (Maidment), ImageJ plug-in (Perez-Ponce) and MIQuaELa (Ayala). Overall agreement better than {approx_equal}90% was found in MTF; the largest differences were observed at frequencies close to the Nyquist limit. For the measurement of NNPS and DQE, agreement is similar to that obtained in the MTF. These results suggest that the developed software can be used with confidence for image quality assessment.

97 MATHEMATICS AND COMPUTING↗

IZI: INFERRING THE GAS PHASE METALLICITY (Z) AND IONIZATION PARAMETER (q) OF IONIZED NEBULAE USING BAYESIAN STATISTICS

We present a new method for inferring the metallicity (Z) and ionization parameter (q) of H II regions and star-forming galaxies using strong nebular emission lines (SELs). We use Bayesian inference to derive the joint and marginalized posterior probability density functions for Z and q given a set of observed line fluxes and an input photoionization model. Our approach allows the use of arbitrary sets of SELs and the inclusion of flux upper limits. The method provides a self-consistent way of determining the physical conditions of ionized nebulae that is not tied to the arbitrary choice of a particular SEL diagnostic and uses all the available information. Unlike theoretically calibrated SEL diagnostics, the method is flexible and not tied to a particular photoionization model. We describe our algorithm, validate it against other methods, and present a tool that implements it called IZI. Using a sample of nearby extragalactic H II regions, we assess the performance of commonly used SEL abundance diagnostics. We also use a sample of 22 local H II regions having both direct and recombination line (RL) oxygen abundance measurements in the literature to study discrepancies in the abundance scale between different methods. We find that oxygen abundances derived through Bayesian inference using currently available photoionization models in the literature can be in good (∼30%) agreement with RL abundances, although some models perform significantly better than others. We also confirm that abundances measured using the direct method are typically ∼0.2 dex lower than both RL and photoionization-model-based abundances.

79 ASTRONOMY AND ASTROPHYSICS↗

Exploring Replica-Exchange Wang-Landau sampling in higher-dimensional parameter space

We considered a higher-dimensional extension for the replica-exchange Wang-Landau algorithm to perform a random walk in the energy and magnetization space of the two-dimensional Ising model. This hybrid scheme combines the advantages of Wang-Landau and Replica-Exchange algorithms, and the one-dimensional version of this approach has been shown to be very efficient and to scale well, up to several thousands of computing cores. This approach allows us to split the parameter space of the system to be simulated into several pieces and still perform a random walk over the entire parameter range, ensuring the ergodicity of the simulation. Previous work, in which a similar scheme of parallel simulation was implemented without using replica exchange and with a different way to combine the result from the pieces, led to discontinuities in the final density of states over the entire range of parameters. From our simulations, it appears that the replica-exchange Wang-Landau algorithm is able to overcome this diculty, allowing exploration of higher parameter phase space by keeping track of the joint density of states.

Valentim, Alexandra↗

Precision Local Burnup Assessment Through Dynamic Peak Fitting in Atom Probe Tomography for Depleted, Enriched, and Irradiated Metallic and Ceramic Fuels

Abstract Burnup estimation in nuclear fuels is vital for evaluating fuel performance, transportation, and safe fuel storage. Accurate assessments of burnup from service period and spent fuels involve tracking the consumption of fissile isotopes of uranium (U) offering a direct insight into energy changes within the fuels especially for thermal spectrum reactors. In current approach, mass spectroscopic technique in atom probe tomography (APT) is utilized for accurate quantification of U isotopes. Quantification of U peaks in mass spectrum is performed on asymmetric shapes due to delayed signals, known as thermal tails, particularly for poorly conducting samples analyzed in laser mode. In this study, we introduce a novel quantification tool for isotopic analysis from APT datasets by developing a fitting algorithm based on shapes of the peaks. A MATLAB-based dynamic peak fitting toolbox is developed and designed to adapt to various peak shapes, ensuring accurate quantification of U isotopes. The effectiveness of this approach is demonstrated in standard Ni-Cr sample, depleted and enriched U samples, and U-based fuels with different burnup levels. The viability of this approach for isotopic quantification is demonstrated on both metallic and ceramic fuels.

Burnup↗

Phase Imaging: A Compressive Sensing Approach

Since Wolfgang Pauli posed the question in 1933, whether the probability densities |Ψ(r)|² (real-space image) and |Ψ(q)|² (reciprocal space image) uniquely determine the wave function Ψ(r) [1], the so called Pauli Problem sparked numerous methods in all fields of microscopy [2, 3]. Reconstructing the complete wave function Ψ(r) = a(r)e-iφ(r) with the amplitude a(r) and the phase φ(r) from the recorded intensity enables the possibility to directly study the electric and magnetic properties of the sample through the phase. In transmission electron microscopy (TEM), electron holography is by far the most established method for phase reconstruction [4]. Requiring a high stability of the microscope, next to the installation of a biprism in the TEM, holography cannot be applied to any microscope straightforwardly. Recently, a phase retrieval approach was proposed using conventional TEM electron diffractive imaging (EDI). Using the SAD aperture as reciprocal-space constraint, a localized sample structure can be reconstructed from its diffraction pattern and a real-space image using the hybrid input-output algorithm [5]. We present an alternative approach using compressive phase-retrieval [6]. Our approach does not require a real-space image. Instead, random complimentary pairs of checkerboard masks are cut into a 200 nm Pt foil covering a conventional TEM aperture (cf. Figure 1). Used as SAD aperture, subsequently diffraction patterns are recorded from the same sample area. Hereby every mask blocks different parts of gold particles on a carbon support (cf. Figure 2). The compressive sensing problem has the following formulation. First, we note that the complex-valued reciprocal-space wave-function is the Fourier transform of the (also complex-valued) real-space wave-function, Ψ(q) = F[Ψ(r)], and subsequently the diffraction pattern image is given by |Ψ(q)|2 = |F[Ψ(r)]|2. We want to find Ψ(r) given a few differently coded diffraction pattern measurements yn = |F[HnΨ(r)]|2, where the matrices Hn encode the mask structure of the aperture. This is a nonlinear inverse problem, but has been shown to be solvable even in the underdetermined case [6]. Since each diffraction pattern yn contains diffraction information from selected regions of the same sample, the differences in each pattern contain local phase information, which can be combined to form a full estimate of the real-space wave-function[7]. References: [1] W. Pauli in “Die allgemeinen Prinzipien der Wellenmechanik“, ed. H Geiger and W Scheel, (Julius Springer, Berlin). [2] A. Tonomura, Rev. Mod. Phys. 59 (1987), p. 639. [3] J. Miao et al, Nature 400 (1999), p. 342. [4] H. Lichte et al, Annu. Rev. Mater. Res. 37 (2007), p. 539. [5] J. Yamasaki et al, Appl. Phys. Lett. 101 (2012), 234105. [6] P Schniter and S Rangan. Signal Proc., IEEE Trans. on. 64(4), (2015), pp. 1043. [7] Supported by the Chemical Imaging, Signature Discovery, and Analytics in Motion initiatives at PNNL. PNNL is operated by Battelle Memorial Inst. for the US DOE; contract DE-AC05-76RL01830.

Schneider, Sebastian↗

Quantifying sampling noise and parametric uncertainty in atomistic-to-continuum simulations using surrogate models

We present a methodology to assess the predictive fidelity of multiscale simulations by incorporating uncertainty in the information exchanged between the components of an atomistic-to-continuum simulation. We account for both the uncertainty due to finite sampling in molecular dynamics (MD) simulations and the uncertainty in the physical parameters of the model. Using Bayesian inference, we represent the expensive atomistic component by a surrogate model that relates the long-term output of the atomistic simulation to its uncertain inputs. We then present algorithms to solve for the variables exchanged across the atomistic-continuum interface in terms of polynomial chaos expansions (PCEs). We also consider a simple Couette flow where velocities are exchanged between the atomistic and continuum components, while accounting for uncertainty in the atomistic model parameters and the continuum boundary conditions. Results show convergence of the coupling algorithm at a reasonable number of iterations. As a result, the uncertainty in the obtained variables significantly depends on the amount of data sampled from the MD simulations and on the width of the time averaging window used in the MD simulations.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Optimizing Long-term Monitoring of Radiation Air Dose Rates - 19169

Radiation air dose rates near the Fukushima Daiichi Nuclear Power Plant (NPP) have been steadily decreasing over the past seven years since the accident in March 2011. After seven years, the radiation monitoring program is expected to transition from short-term emergency responses to long-term monitoring. The objectives of long-term monitoring are often different from early emergency responses, since extensive data accumulation has led to an understanding of contaminant distributions and mobility. The main long-term monitoring objectives are (1) to confirm system stability and the continuing reduction of contaminant and hazard levels, (2) to provide assurance for the public, (3) to accumulate the basic datasets for scientific knowledge and future preparation, and (4) to detect changes or anomalies in contaminant mobility (if they occur), or any unexpected processes or events. In this work, we have developed a methodology for optimizing the sampling locations of radiation air dose-rate monitoring. Three steps are taken in order to determine sampling locations in a systematic manner: (1) prioritizing the critical locations, such as schools or regulatory requirement locations, (2) diversifying locations across the key environmental controls that are known to influence contaminant mobility and distributions, and (3) capturing the heterogeneity of radiation air dose rates across the domain. In the second step, we use a Gaussian mixture model to identify the representative locations in multiple environmental variables, such as elevation and land-cover types. In the third step, we use a Gaussian process model to capture and estimate the heterogeneity of air dose rates across the domain. We use the integrated dose-rate map from the Bayesian geostatistical method as the reference map and distribute the sampling in such a way as to capture the heterogeneity of the reference map. Although we assume hypothetical sets of measurement locations to evaluate the performance of the methodology and algorithms, we have demonstrated this approach using real, spatially extensive datasets, including the integrated air-dose-rate map of 2016. Our results have shown that this approach allows us to add sampling locations in a systematic manner. Increasing the number of sampling locations can better capture the heterogeneity of dose rates, although the estimation error does not decrease further after a certain number of samples. We have also found that when there are restrictions such as pre-existing monitoring locations or the ones along roads, the spatial estimation becomes poor even with the same number of monitoring locations. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

A PUBLIC CATALOG OF STELLAR MASSES, STAR FORMATION AND METALLICITY HISTORIES, AND DUST CONTENT FROM THE SLOAN DIGITAL SKY SURVEY USING VESPA

We applied the VESPA algorithm to the Sloan Digital Sky Survey final data release of the Main Galaxies and Luminous Red Galaxies samples. The result is a catalog of stellar masses, detailed star formation and metallicity histories and dust content of nearly 800,000 galaxies. We make the catalog public via a T-SQL database, which is described in detail in this paper. We present the results using a range of stellar population and dust models, and will continue to update the catalog as new and improved models are made public. We also present a brief exploration of the catalog, and show that the quantities derived are robust: luminous red galaxies can be described by one to three populations, whereas a main galaxy sample galaxy needs on average two to five; red galaxies are older and less dusty; the dust values we recover are well correlated with measured Balmer decrements and star formation rates are also in agreement with previous measurements. We find that whereas some derived quantities are robust to the choice of modelling, many are still not.

79 ASTRONOMY AND ASTROPHYSICS↗

REMOVING BIASES IN RESOLVED STELLAR MASS MAPS OF GALAXY DISKS THROUGH SUCCESSIVE BAYESIAN MARGINALIZATION

Stellar masses of galaxies are frequently obtained by fitting stellar population synthesis models to galaxy photometry or spectra. The state of the art method resolves spatial structures within a galaxy to assess the total stellar mass content. In comparison to unresolved studies, resolved methods yield, on average, higher fractions of stellar mass for galaxies. In this work we improve the current method in order to mitigate a bias related to the resolved spatial distribution derived for the mass. The bias consists in an apparent filamentary mass distribution and a spatial coincidence between mass structures and dust lanes near spiral arms. The improved method is based on iterative Bayesian marginalization, through a new algorithm we have named Bayesian Successive Priors (BSP). We have applied BSP to M51 and to a pilot sample of 90 spiral galaxies from the Ohio State University Bright Spiral Galaxy Survey. By quantitatively comparing both methods, we find that the average fraction of stellar mass missed by unresolved studies is only half what previously thought. In contrast with the previous method, the output BSP mass maps bear a better resemblance to near-infrared images.

79 ASTRONOMY AND ASTROPHYSICS↗

Deep probabilistic direction prediction in 3D with applications to directional dark matter detectors

Abstract We present the first method to probabilistically predict 3D direction in a deep neural network model. The probabilistic predictions are modeled as a heteroscedastic von Mises-Fisher distribution on the sphere S 2 , giving a simple way to quantify aleatoric uncertainty. This approach generalizes the cosine distance loss which is a special case of our loss function when the uncertainty is assumed to be uniform across samples. We develop approximations required to make the likelihood function and gradient calculations stable. The method is applied to the task of predicting the 3D directions of electrons, the most complex signal in a class of experimental particle physics detectors designed to demonstrate the particle nature of dark matter and study solar neutrinos. Using simulated Monte Carlo data, the initial direction of recoiling electrons is inferred from their tortuous trajectories, as captured by the 3D detectors. For 40 keV electrons in a 70% He 30% CO 2 gas mixture at STP, the new approach achieves a mean cosine distance of 0.104 (26 ∘ ) compared to 0.556 (64 ∘ ) achieved by a non-machine learning algorithm. We show that the model is well-calibrated and accuracy can be increased further by removing samples with high predicted uncertainty. This advancement in probabilistic 3D directional learning could increase the sensitivity of directional dark matter detectors.

Computer Science↗