-
Lessons Learned from the Two Largest Galaxy Morphological Classification Catalogues built by Convolutional Neural Networks
Authors:
Ting-Yun Cheng,
H. Domínguez Sánchez,
J. Vega-Ferrero,
C. J. Conselice,
M. Siudek,
A. Aragón-Salamanca,
M. Bernardi,
R. Cooke,
L. Ferreira,
M. Huertas-Company,
J. Krywult,
A. Palmese,
A. Pieres,
A. A. Plazas Malagón,
A. Carnero Rosell,
D. Gruen,
D. Thomas,
D. Bacon,
D. Brooks,
D. J. James,
D. L. Hollowood,
D. Friedel,
E. Suchyta,
E. Sanchez,
F. Menanteau
, et al. (32 additional authors not shown)
Abstract:
We compare the two largest galaxy morphology catalogues, which separate early and late type galaxies at intermediate redshift. The two catalogues were built by applying supervised deep learning (convolutional neural networks, CNNs) to the Dark Energy Survey data down to a magnitude limit of $\sim$21 mag. The methodologies used for the construction of the catalogues include differences such as the…
▽ More
We compare the two largest galaxy morphology catalogues, which separate early and late type galaxies at intermediate redshift. The two catalogues were built by applying supervised deep learning (convolutional neural networks, CNNs) to the Dark Energy Survey data down to a magnitude limit of $\sim$21 mag. The methodologies used for the construction of the catalogues include differences such as the cutout sizes, the labels used for training, and the input to the CNN - monochromatic images versus $gri$-band normalized images. In addition, one catalogue is trained using bright galaxies observed with DES ($i<18$), while the other is trained with bright galaxies ($r<17.5$) and `emulated' galaxies up to $r$-band magnitude $22.5$. Despite the different approaches, the agreement between the two catalogues is excellent up to $i<19$, demonstrating that CNN predictions are reliable for samples at least one magnitude fainter than the training sample limit. It also shows that morphological classifications based on monochromatic images are comparable to those based on $gri$-band images, at least in the bright regime. At fainter magnitudes, $i>19$, the overall agreement is good ($\sim$95\%), but is mostly driven by the large spiral fraction in the two catalogues. In contrast, the agreement within the elliptical population is not as good, especially at faint magnitudes. By studying the mismatched cases we are able to identify lenticular galaxies (at least up to $i<19$), which are difficult to distinguish using standard classification approaches. The synergy of both catalogues provides an unique opportunity to select a population of unusual galaxies.
△ Less
Submitted 14 September, 2022;
originally announced September 2022.
-
Moiré-localized interlayer exciton wavefunctions captured by imaging its electron and hole constituents
Authors:
Ouri Karni,
Elyse Barré,
Vivek Pareek,
Johnathan D. Georgaras,
Michael K. L. Man,
Chakradhar Sahoo,
David R. Bacon,
Xing Zhu,
Henrique B. Ribeiro,
Aidan L. O'Beirne,
Jenny Hu,
Abdullah Al-Mahboob,
Mohamed M. M. Abdelrasoul,
Nicholas S. Chan,
Arka Karmakar,
Andrew J. Winchester,
Bumho Kim,
Kenji Watanabe,
Takashi Taniguchi,
Katayun Barmak,
Julien Madéo,
Felipe H. da Jornada,
Tony F. Heinz,
Keshav M. Dani
Abstract:
Interlayer excitons (ILXs) - electron-hole pairs bound across two atomically thin layered semiconductors - have emerged as attractive platforms to study exciton condensation, single-photon emission and other quantum-information applications. Yet, despite extensive optical spectroscopic investigations, critical information about their size, valley configuration and the influence of the moiré potent…
▽ More
Interlayer excitons (ILXs) - electron-hole pairs bound across two atomically thin layered semiconductors - have emerged as attractive platforms to study exciton condensation, single-photon emission and other quantum-information applications. Yet, despite extensive optical spectroscopic investigations, critical information about their size, valley configuration and the influence of the moiré potential remains unknown. Here, we captured images of the time- and momentum-resolved distribution of both the electron and the hole that bind to form the ILX in a WSe2/MoS2 heterostructure. We thereby obtain a direct measurement of the interlayer exciton diameter of ~5.4 nm, comparable to the moiré unit-cell length of 6.1 nm. Surprisingly, this large ILX is well localized within the moiré cell to a region of only 1.8 nm - smaller than the size of the exciton itself. This high degree of localization of the interlayer exciton is backed by Bethe-Salpeter equation calculations and demonstrates that the ILX can be localized within small moiré unit cells. Unlike large moiré cells, these are uniform over large regions, thus allowing the formation of extended arrays of localized excitations for quantum technology.
△ Less
Submitted 4 August, 2021;
originally announced August 2021.
-
Investigating Cosmological GAN Emulators Using Latent Space Interpolation
Authors:
Andrius Tamosiunas,
Hans A. Winther,
Kazuya Koyama,
David J. Bacon,
Robert C. Nichol,
Ben Mawdsley
Abstract:
Generative adversarial networks (GANs) have been recently applied as a novel emulation technique for large scale structure simulations. Recent results show that GANs can be used as a fast, efficient and computationally cheap emulator for producing novel weak lensing convergence maps as well as cosmic web data in 2-D and 3-D. However, like any algorithm, the GAN approach comes with a set of limitat…
▽ More
Generative adversarial networks (GANs) have been recently applied as a novel emulation technique for large scale structure simulations. Recent results show that GANs can be used as a fast, efficient and computationally cheap emulator for producing novel weak lensing convergence maps as well as cosmic web data in 2-D and 3-D. However, like any algorithm, the GAN approach comes with a set of limitations, such as an unstable training procedure and the inherent randomness of the produced outputs. In this work we employ a number of techniques commonly used in the machine learning literature to address the mentioned limitations. In particular, we train a GAN to produce both weak lensing convergence maps and dark matter overdensity field data for multiple redshifts, cosmological parameters and modified gravity models. In addition, we train a GAN using the newest Illustris data to emulate dark matter, gas and internal energy distribution data simultaneously. Finally, we apply the technique of latent space interpolation to control which outputs the algorithm produces. Our results indicate a 1-20% difference between the power spectra of the GAN-produced and the training data samples depending on the dataset used and whether Gaussian smoothing was applied. Finally, recent research on generative models suggests that such algorithms can be treated as map**s from a lower-dimensional input (latent) space to a higher dimensional (data) manifold. We explore such a theoretical description as a tool for better understanding the latent space interpolation procedure.
△ Less
Submitted 29 June, 2021; v1 submitted 21 April, 2020;
originally announced April 2020.
-
Hartree-Fock on a superconducting qubit quantum computer
Authors:
Frank Arute,
Kunal Arya,
Ryan Babbush,
Dave Bacon,
Joseph C. Bardin,
Rami Barends,
Sergio Boixo,
Michael Broughton,
Bob B. Buckley,
David A. Buell,
Brian Burkett,
Nicholas Bushnell,
Yu Chen,
Zijun Chen,
Benjamin Chiaro,
Roberto Collins,
William Courtney,
Sean Demura,
Andrew Dunsworth,
Daniel Eppens,
Edward Farhi,
Austin Fowler,
Brooks Foxen,
Craig Gidney,
Marissa Giustina
, et al. (57 additional authors not shown)
Abstract:
As the search continues for useful applications of noisy intermediate scale quantum devices, variational simulations of fermionic systems remain one of the most promising directions. Here, we perform a series of quantum simulations of chemistry the largest of which involved a dozen qubits, 78 two-qubit gates, and 114 one-qubit gates. We model the binding energy of ${\rm H}_6$, ${\rm H}_8$,…
▽ More
As the search continues for useful applications of noisy intermediate scale quantum devices, variational simulations of fermionic systems remain one of the most promising directions. Here, we perform a series of quantum simulations of chemistry the largest of which involved a dozen qubits, 78 two-qubit gates, and 114 one-qubit gates. We model the binding energy of ${\rm H}_6$, ${\rm H}_8$, ${\rm H}_{10}$ and ${\rm H}_{12}$ chains as well as the isomerization of diazene. We also demonstrate error-mitigation strategies based on $N$-representability which dramatically improve the effective fidelity of our experiments. Our parameterized ansatz circuits realize the Givens rotation approach to non-interacting fermion evolution, which we variationally optimize to prepare the Hartree-Fock wavefunction. This ubiquitous algorithmic primitive corresponds to a rotation of the orbital basis and is required by many proposals for correlated simulations of molecules and Hubbard models. Because non-interacting fermion evolutions are classically tractable to simulate, yet still generate highly entangled states over the computational basis, we use these experiments to benchmark the performance of our hardware while establishing a foundation for scaling up more complex correlated quantum simulations of chemistry.
△ Less
Submitted 18 September, 2020; v1 submitted 8 April, 2020;
originally announced April 2020.
-
Verifying Spent Fuel Containers Before Deep Geological Storage with Cosmic Ray Muons
Authors:
D. Poulson,
J. M. Durham,
J. D. Bacon,
E. Guardincerri,
C. L. Morris
Abstract:
International nuclear safeguards inspectors do not have a method to verify the contents of sealed storage casks containing spent reactor fuel. The heavy shielding that is used to limit radiation emission attenuates and scatters photons and neutrons emitted by the fuel, and thereby hinders inspection with these probes. This problem is especially pressing given the policy decisions of several nation…
▽ More
International nuclear safeguards inspectors do not have a method to verify the contents of sealed storage casks containing spent reactor fuel. The heavy shielding that is used to limit radiation emission attenuates and scatters photons and neutrons emitted by the fuel, and thereby hinders inspection with these probes. This problem is especially pressing given the policy decisions of several nations to begin permanent disposal of spent fuel in deep geological repositories. Radiography with cosmic-ray muons provides a potential solution, as muons are able to penetrate the cask and fuel and provide information on the cask contents. Here we show in simulation that muon scattering radiography can be used to inspect the contents of sealed geological storage casks, and can discern between a variety of plausible diversion scenarios. This technique can be applied immediately prior to permanent interment in a geological repository, giving inspectors a final opportunity to verify State declarations of spent fuel disposal.
△ Less
Submitted 9 May, 2019;
originally announced May 2019.
-
Learning from the machine: interpreting machine learning algorithms for point- and extended- source classification
Authors:
Xan Morice-Atkinson,
Ben Hoyle,
David Bacon
Abstract:
We investigate star-galaxy classification for astronomical surveys in the context of four methods enabling the interpretation of black-box machine learning systems. The first is outputting and exploring the decision boundaries as given by decision tree based methods, which enables the visualization of the classification categories. Secondly, we investigate how the Mutual Information based Transduc…
▽ More
We investigate star-galaxy classification for astronomical surveys in the context of four methods enabling the interpretation of black-box machine learning systems. The first is outputting and exploring the decision boundaries as given by decision tree based methods, which enables the visualization of the classification categories. Secondly, we investigate how the Mutual Information based Transductive Feature Selection (MINT) algorithm can be used to perform feature pre-selection. If one would like to provide only a small number of input features to a machine learning classification algorithm, feature pre-selection provides a method to determine which of the many possible input properties should be selected. Third is the use of the tree-interpreter package to enable popular decision tree based ensemble methods to be opened, visualized, and understood. This is done by additional analysis of the tree based model, determining not only which features are important to the model, but how important a feature is for a particular classification given its value. Lastly, we use decision boundaries from the model to revise an already existing method of classification, essentially asking the tree based method where decision boundaries are best placed and defining a new classification method.
We showcase these techniques by applying them to the problem of star-galaxy separation using data from the Sloan Digital Sky Survey (hereafter SDSS). We use the output of MINT and the ensemble methods to demonstrate how more complex decision boundaries improve star-galaxy classification accuracy over the standard SDSS frames approach (reducing misclassifications by up to $\approx33\%$). We then show how tree-interpreter can be used to explore how relevant each photometric feature is when making a classification on an object by object basis.
△ Less
Submitted 11 December, 2017;
originally announced December 2017.
-
Cosmic ray muon computed tomography of spent nuclear fuel in dry storage casks
Authors:
D. Poulson,
J. M. Durham,
E. Guardincerri,
C. L. Morris,
J. D. Bacon,
K. Plaud-Ramos,
D. Morley,
A. Hecht
Abstract:
Radiography with cosmic ray muon scattering has proven to be a successful method of imaging nuclear material through heavy shielding. Of particular interest is monitoring dry storage casks for diversion of plutonium contained in spent reactor fuel. Using muon tracking detectors that surround a cylindrical cask, cosmic ray muon scattering can be simultaneously measured from all azimuthal angles, gi…
▽ More
Radiography with cosmic ray muon scattering has proven to be a successful method of imaging nuclear material through heavy shielding. Of particular interest is monitoring dry storage casks for diversion of plutonium contained in spent reactor fuel. Using muon tracking detectors that surround a cylindrical cask, cosmic ray muon scattering can be simultaneously measured from all azimuthal angles, giving complete tomographic coverage of the cask interior. This paper describes the first application of filtered back projection algorithms, typically used in medical imaging, to cosmic ray muon imaging. The specific application to monitoring spent nuclear fuel in dry storage casks is investigated via GEANT4 simulations. With a cylindrical muon tracking detector surrounding a typical spent fuel cask, the cask contents can be confirmed with high confidence in less than two days exposure. Similar results can be obtained by moving a smaller detector to view the cask from multiple angles.
△ Less
Submitted 29 April, 2016;
originally announced April 2016.
-
Imaging the inside of thick structures using cosmic rays
Authors:
E. Guardincerri,
J. M. Durham,
C. Morris,
J. D. Bacon,
T. M. Daughton,
S. Fellows,
O. R. Johnson,
D. J. Morley,
K. Plaud-Ramos,
D. C. Poulson,
Z. Wang
Abstract:
The authors present here a new method to image reinforcement elements inside thick structures and the results of a demonstration measurement performed on a mock-up wall built at Los Alamos National Laboratory. The method, referred to as "multiple scattering muon radiography", relies on the use of cosmic-ray muons as probes. The work described in this article was performed to prove the viability of…
▽ More
The authors present here a new method to image reinforcement elements inside thick structures and the results of a demonstration measurement performed on a mock-up wall built at Los Alamos National Laboratory. The method, referred to as "multiple scattering muon radiography", relies on the use of cosmic-ray muons as probes. The work described in this article was performed to prove the viability of the technique as a means to image the interior of the dome of Florence Cathedral Santa Maria del Fiore, one of the UNESCO World Heritage sites and among the highest profile buildings in existence. Its result shows the effectiveness of the technique as a tool to radiograph thick structures and image denser object inside them.
△ Less
Submitted 6 January, 2016;
originally announced January 2016.
-
A multilayer surface detector for ultracold neutrons
Authors:
Zhehui Wang,
M. A. Hoffbauer,
C. L. Morris,
N. B. Callahan,
E. R. Adamek,
J. D. Bacon,
M. Blatnik,
A. E. Brandt,
L. J. Broussard,
S. M. Clayton,
C. Cude-Woods,
S. Currie,
E. B. Dees,
X. Ding,
J. Gao,
F. E. Gray,
K. P. Hickerson,
A. T. Holley,
T. M. Ito,
C. -Y. Liu,
M. Makela,
J. C. Ramsey,
R. W. Pattie, Jr.,
D. J. Salvat,
A. Saunders
, et al. (11 additional authors not shown)
Abstract:
A multilayer surface detector for ultracold neutrons (UCNs) is described. The top $^{10}$B layer is exposed to the vacuum chamber and directly captures UCNs. The ZnS:Ag layer beneath the $^{10}$B layer is a few microns thick, which is sufficient to detect the charged particles from the $^{10}$B(n,$α$)$^7$Li neutron-capture reaction, while thin enough so that ample light due to $α$ and $^7$Li escap…
▽ More
A multilayer surface detector for ultracold neutrons (UCNs) is described. The top $^{10}$B layer is exposed to the vacuum chamber and directly captures UCNs. The ZnS:Ag layer beneath the $^{10}$B layer is a few microns thick, which is sufficient to detect the charged particles from the $^{10}$B(n,$α$)$^7$Li neutron-capture reaction, while thin enough so that ample light due to $α$ and $^7$Li escapes for detection by photomultiplier tubes. One-hundred-nm thick $^{10}$B layer gives high UCN detection efficiency, as determined by the mean UCN kinetic energy, detector materials and others. Low background, including negligible sensitivity to ambient neutrons, has also been verified through pulse-shape analysis and comparisons with other existing $^3$He and $^{10}$B detectors. This type of detector has been configured in different ways for UCN flux monitoring, development of UCN guides and neutron lifetime research.
△ Less
Submitted 24 April, 2015; v1 submitted 11 March, 2015;
originally announced March 2015.
-
A double-helix neutron detector using micron-size B-10 powder
Authors:
Zhehui Wang,
C. L. Morris,
J. D. Bacon,
M. I. Brockwell,
J. C. Ramsey
Abstract:
A double-helix electrode configuration is combined with a $^{10}$B powder coating technique to build large-area (9 in $\times$ 36 in) neutron detectors. The neutron detection efficiency for each of the four prototypes is comparable to a single 2-bar $^3$He drift tube of the same length (36 in). One unit has been operational continuously for 18 months and the change of efficiency is less than 1%. A…
▽ More
A double-helix electrode configuration is combined with a $^{10}$B powder coating technique to build large-area (9 in $\times$ 36 in) neutron detectors. The neutron detection efficiency for each of the four prototypes is comparable to a single 2-bar $^3$He drift tube of the same length (36 in). One unit has been operational continuously for 18 months and the change of efficiency is less than 1%. An analytic model for pulse heigh spectra is described and the predicted mean film thickness agrees with the experiment to within 30%. Further detector optimization is possible through film texture, power size, moderator box and gas. The estimated production cost per unit is less than 3k US\$ and the technology is thus suitable for deployment in large numbers.
△ Less
Submitted 19 June, 2014;
originally announced June 2014.
-
Detecting Special Nuclear Material Using Muon-Induced Neutron Emission
Authors:
E. Guardincerri,
J. D. Bacon,
K. Borodzin,
J. M. Durham,
J. M. Fabritius II,
A. Hecht,
E. C. Milner,
H. Miyadera,
C. L. Morris,
J. O. Perry,
D. Poulson
Abstract:
The penetrating ability of cosmic ray muons makes them an attractive probe for imaging dense materials. Here, we describe experimental results from a new technique that uses neutrons generated by cosmic-ray muons to identify the presence of special nuclear material (SNM). Neutrons emitted from SNM are used to tag muon-induced fission events in actinides and laminography is used to form images of t…
▽ More
The penetrating ability of cosmic ray muons makes them an attractive probe for imaging dense materials. Here, we describe experimental results from a new technique that uses neutrons generated by cosmic-ray muons to identify the presence of special nuclear material (SNM). Neutrons emitted from SNM are used to tag muon-induced fission events in actinides and laminography is used to form images of the stop** material. This technique allows the imaging of SNM-bearing objects tagged using muon tracking detectors located above or to the side of the objects, and may have potential applications in warhead verification scenarios. During the experiment described here we did not attempt to distinguish the type or grade of the SNM.
△ Less
Submitted 25 March, 2015; v1 submitted 4 June, 2014;
originally announced June 2014.
