arXiv subjects

Yang Sun

Publications and source records attributed to Yang Sun.

At least 181 records · Page 10Linked to original sources

arXiv2021

Non-Convex Tensor Low-Rank Approximation for Infrared Small Target Detection

Infrared small target detection is an important fundamental task in the infrared system. Therefore, many infrared small target detection methods have been proposed, in which the low-rank model has been used as a powerful tool. However, most low-rank-based methods assign the same weights for different singular values, which will lead to inaccurate background estimation. Considering that different singular values have different importance and should be treated discriminatively, in this paper, we propose a non-convex tensor low-rank approximation (NTLA) method for infrared small target detection. In our method, NTLA regularization adaptively assigns different weights to different singular values for accurate background estimation. Based on the proposed NTLA, we propose asymmetric spatial-temporal total variation (ASTTV) regularization to achieve more accurate background estimation in complex scenes. Compared with the traditional total variation approach, ASTTV exploits different smoothness intensities for spatial and temporal regularization. We design an efficient algorithm to find the optimal solution of our method. Compared with some state-of-the-art methods, the proposed method achieves an improvement in terms of various evaluation metrics. Extensive experimental results in various complex scenes demonstrate that our method has strong robustness and low false-alarm rate. Code is available at https://github.com/LiuTing20a/ASTTV-NTLA.

cs.CV↗

arXiv2021

Iron-rich Fe-O compounds with closest-packed layers at core pressures

Oxygen solubility in solid iron is extremely low, even at high pressures and temperatures. Thus far, no Fe-O compounds between Fe and FeO endmembers have been reported experimentally. We observed chemical reactions of Fe with FeO or Fe$_2$O$_3$ $in\ situ$ x-ray diffraction experiments at 220-260 GPa and 3,000-3,500 K. The refined diffraction patterns are consistent with a series of Fe$_n$O (n $>$ 1) compounds (e.g., Fe$_{25}$O$_{13}$ and Fe$_{28}$O$_{14}$) identified using the adaptive genetic algorithm. Like $ε$-Fe in the hexagonal close-packed (hcp) structure, the structures of Fe$_n$O compounds consist of oxygen-only close-packed monolayers distributed between iron-only layers. $Ab\ initio$ calculations show systematic electronic properties of these compounds that have ramifications for the physical properties of Earth's inner core.

cond-mat.mtrl-sci↗

arXiv2021

StackRec: Efficient Training of Very Deep Sequential Recommender Models by Iterative Stacking

Deep learning has brought great progress for the sequential recommendation (SR) tasks. With advanced network architectures, sequential recommender models can be stacked with many hidden layers, e.g., up to 100 layers on real-world recommendation datasets. Training such a deep network is difficult because it can be computationally very expensive and takes much longer time, especially in situations where there are tens of billions of user-item interactions. To deal with such a challenge, we present StackRec, a simple, yet very effective and efficient training framework for deep SR models by iterative layer stacking. Specifically, we first offer an important insight that hidden layers/blocks in a well-trained deep SR model have very similar distributions. Enlightened by this, we propose the stacking operation on the pre-trained layers/blocks to transfer knowledge from a shallower model to a deep model, then we perform iterative stacking so as to yield a much deeper but easier-to-train SR model. We validate the performance of StackRec by instantiating it with four state-of-the-art SR models in three practical scenarios with real-world datasets. Extensive experiments show that StackRec achieves not only comparable performance, but also substantial acceleration in training time, compared to SR models that are trained from scratch. Codes are available at https://github.com/wangjiachun0426/StackRec.

cs.IR↗

arXiv2021

Unconventional iron-magnesium compounds at terapascal pressures

Being a lithophile element at ambient pressure, magnesium is long believed to be immiscible with iron. A recent study by Gao et al. [1] showed that pressure turns magnesium into a siderophile element and can produce unconventional Fe-Mg compounds. Here, we extend the investigation to exoplanetary pressure conditions using an adaptive genetic algorithm-based variable-composition structural prediction approach. We identify several Fe-Mg phases up to 3 TPa. Our cluster alignment analysis reveals that most of the predicted Fe-Mg compounds prefer a BCC packing motif at terapascal pressures. This study provides a more comprehensive structure database to support future investigations of the high-pressure structural behavior of Fe-Mg and ternary, quaternary, etc. compounds involving these elements.

cond-mat.mtrl-sci↗

arXiv2021

$FM^2$: Field-matrixed Factorization Machines for Recommender Systems

Click-through rate (CTR) prediction plays a critical role in recommender systems and online advertising. The data used in these applications are multi-field categorical data, where each feature belongs to one field. Field information is proved to be important and there are several works considering fields in their models. In this paper, we proposed a novel approach to model the field information effectively and efficiently. The proposed approach is a direct improvement of FwFM, and is named as Field-matrixed Factorization Machines (FmFM, or $FM^2$). We also proposed a new explanation of FM and FwFM within the FmFM framework, and compared it with the FFM. Besides pruning the cross terms, our model supports field-specific variable dimensions of embedding vectors, which acts as soft pruning. We also proposed an efficient way to minimize the dimension while keeping the model performance. The FmFM model can also be optimized further by caching the intermediate vectors, and it only takes thousands of floating-point operations (FLOPs) to make a prediction. Our experiment results show that it can out-perform the FFM, which is more complex. The FmFM model's performance is also comparable to DNN models which require much more FLOPs in runtime.

cs.IR↗

arXiv2021

Sensitivity of Neutron-Rich Nuclear Isomer Behavior to Uncertainties in Direct Transitions

Nuclear isomers are populated in the rapid neutron capture process (r process) of nucleosynthesis. The r process may cover a wide range of temperatures, potentially starting from several tens of GK (several MeV) and then cooling as material is ejected from the event. As the r-process environment cools, isomers can freeze out of thermal equilibrium or be directly populated as astrophysical isomers (astromers). Two key behaviors of astromers -- ground state<->isomer transition rates and thermalization temperatures -- are determined by direct transition rates between pairs of nuclear states. We perform a sensitivity study to constrain the effects of unknown transitions on astromer behavior. We also introduce a categorization of astromers that describes their potential effects in hot environments. We provide a table of neutron-rich isomers that includes the astromer type, thermalization temperature, and key unmeasured transition rates.

astro-ph.HE↗

arXiv2021

Structural Interventions in Networks

Two types of interventions are commonly implemented in networks: characteristic intervention, which influences individuals' intrinsic incentives, and structural intervention, which targets the social links among individuals. In this paper we provide a general framework to evaluate the distinct equilibrium effects of both types of interventions. We identify a hidden equivalence between a structural intervention and an endogenously determined characteristic intervention. Compared with existing approaches in the literature, the perspective from such an equivalence provides several advantages in the analysis of interventions that target network structure. We present a wide range of applications of our theory, including identifying the most wanted criminal(s) in delinquent networks and targeting the key connector for isolated communities.

econ.TH↗

arXiv2021

Urca Cooling in Neutron Star Crusts and Oceans: Effects of Nuclear Excitations

The excited-state structure of atomic nuclei can modify nuclear processes in stellar environments. In this work, we study the influence of nuclear excitations on Urca cooling (repeated back-and-forth beta decay and electron capture in a pair of nuclear isotopes) in the crust and ocean of neutron stars. We provide for the first time an expression for Urca process neutrino luminosity which accounts for excited states of both members of an Urca pair. We use our new formula with state-of-the-art nuclear structure inputs to compute neutrino luminosities of candidate Urca cooling pairs. Our nuclear inputs consist of the latest experimental data supplemented with calculations using the projected shell model. We show that, in contrast with previous results that only consider the ground states of both nuclei in the pair, our calculated neutrino luminosities for different Urca pairs vary sensitively with the environment temperature and can be radically different from those obtained in the one transition approximation.

nucl-th↗

arXiv2021

Prediction of crystal structures and motifs in the Fe-Mg-O system at Earth's core pressures

Fe, Mg, and O are among the most abundant elements in terrestrial planets. While the behavior of the Fe-O, Mg-O, and Fe-Mg binary systems under pressure have been investigated, there are still very few studies of the Fe-Mg-O ternary system at relevant Earth's core and super-Earth's mantle pressures. Here, we use the adaptive genetic algorithm (AGA) to study ternary Fe$_x$Mg$_y$O$_z$ phases in a wide range of stoichiometries at 200 GPa and 350 GPa. We discovered three dynamically stable phases with stoichiometries FeMg$_2$O$_4$, Fe$_2$MgO$_4$, and FeMg$_3$O$_4$ with lower enthalpy than any known combination of Fe-Mg-O high-pressure compounds at 350 GPa. With the discovery of these phases, we construct the Fe-Mg-O ternary convex hull. We further clarify the composition- and pressure-dependence of structural motifs with the analysis of the AGA-found stable and metastable structures. Analysis of binary and ternary stable phases suggest that O, Mg, or both could stabilize a BCC iron alloy at inner core pressures.

cond-mat.mtrl-sci↗

arXiv2020

Spatial decomposition of magnetic anisotropy in magnets: application for doped Fe16N2

We propose a scheme of decomposition of the total relativistic energy in solids to intra- and interatomic contributions. The method is based on a variation of the speed of light from its value in relativistic theory to infinity (a non-relativistic limit). As an illustration of the method, we tested such decomposition in the case of a spin-orbit interaction variation for decomposition of the magnetic anisotropy energy (MAE) in CoPt. We further studied the α''-Fe16N2 magnet doped by Bi, Sb, Co and Pt atoms. It has been found that the addition of Pt atoms can enhance the MAE by as large as five times while Bi and Sb substitutions double the total MAE. Using the proposed technique we demonstrate the spatial distribution of these enhancements. Our studies also suggest that Sb, Pt and Co substitutions could be synthesized by experiments.

cond-mat.mtrl-sci↗

arXiv2020

Astromers: Nuclear Isomers in Astrophysics

We develop a method to compute thermally-mediated transition rates between the ground state and long-lived isomers in nuclei. We also establish criteria delimiting a thermalization temperature above which a nucleus may be considered a single species and below which it must be treated as two separate species: a ground state species, and an astrophysical isomer ("astromer") species. Below the thermalization temperature, the destruction rates dominate the internal transition rates between the ground state and the isomer. If the destruction rates also differ greatly from one another, the nuclear levels fall out of or fail to reach thermal equilibrium. Without thermal equilibrium, there may not be a safe assumption about the distribution of occupation probability among the nuclear levels when computing nuclear reaction rates. In these conditions, the isomer has astrophysical consequences and should be treated a separate astromer species which evolves separately from the ground state in a nucleosynthesis network. We apply our transition rate methods and perform sensitivity studies on a few well-known astromers. We also study transitions in several other isomers of likely astrophysical interest.

astro-ph.HE↗

arXiv2020

A Star-based Method for Precise Wavelength Calibration of the Chinese Space Station Telescope (CSST) Slitless Spectroscopic Survey

The Chinese Space Station Telescope (CSST) spectroscopic survey aims to deliver high-quality low-resolution ($R > 200$) slitless spectra for hundreds of millions of targets down to a limiting magnitude of about 21 mag, distributed within a large survey area (17500 deg$^2$) and covering a wide wavelength range (255-1000 nm by 3 bands GU, GV, and GI). As slitless spectroscopy precludes the usage of wavelength calibration lamps, wavelength calibration is one of the most challenging issues in the reduction of slitless spectra, yet it plays a key role in measuring precise radial velocities of stars and redshifts of galaxies. In this work, we propose a star-based method that can monitor and correct for possible errors in the CSST wavelength calibration using normal scientific observations, taking advantage of the facts that i) there are about ten million stars with reliable radial velocities now available thanks to spectroscopic surveys like LAMOST, ii) the large field of view of CSST enables efficient observations of such stars in a short period of time, and iii) radial velocities of such stars can be reliably measured using only a narrow segment of CSST spectra. We demonstrate that it is possible to achieve a wavelength calibration precision of a few $\mathrm{km}\,\mathrm{s}^{-1}$ for the GU band, and about 10 to 20 $\mathrm{km}\,\mathrm{s}^{-1}$ for the GV and GI bands, with only a few hundred velocity standard stars. Implementations of the method to other surveys are also discussed.

astro-ph.IM↗

arXiv2020

Precision of The Chinese Space Station Telescope (CSST) Stellar Radial Velocities

The Chinese Space Station Telescope (CSST) spectroscopic survey plans to deliver high-quality low-resolution ($R > 200$) slitless spectra for hundreds of millions of targets down to a limiting magnitude of about 21 mag, covering a large survey area (17500 deg$^2$) and a wide wavelength range (255-1000 nm by 3 bands GU, GV, and GI). In this work, we use empirical spectra of the Next Generation Spectral Library to simulate the CSST stellar spectra at $R = 250$, and investigate their capabilities in measuring radial velocities. We find that velocity uncertainties depend strongly on effective temperature, weakly on metallicity for only FGK stars, and hardly on surface gravity. It is possible to deliver stellar radial velocities to a precision of about $3 \,\mathrm{km}\,\mathrm{s}^{-1}$ for AFGKM stars, and about $10 \,\mathrm{km}\,\mathrm{s}^{-1}$ for OB stars, at signal-to-noise ratio (SNR) of 100. Velocity uncertainties using single GU/GV/GI band spectra are also explored. Given the same SNR, the GU band performs best, the GV band the second best, and then the GI band. The effects of spectral normalization and imperfect template on velocity measurements are investigated and found to be very weak. The uncertainties caused by wavelength calibration are considered and found to be moderate. Given the possible precision of radial velocities, the CSST spectroscopic survey can enable interesting science such as searching for hyper-velocity stars. Limitations of our results are also discussed.

astro-ph.IM↗

arXiv2020

MPG-Net: Multi-Prediction Guided Network for Segmentation of Retinal Layers in OCT Images

Optical coherence tomography (OCT) is a commonly-used method of extracting high resolution retinal information. Moreover there is an increasing demand for the automated retinal layer segmentation which facilitates the retinal disease diagnosis. In this paper, we propose a novel multiprediction guided attention network (MPG-Net) for automated retinal layer segmentation in OCT images. The proposed method consists of two major steps to strengthen the discriminative power of a U-shape Fully convolutional network (FCN) for reliable automated segmentation. Firstly, the feature refinement module which adaptively re-weights the feature channels is exploited in the encoder to capture more informative features and discard information in irrelevant regions. Furthermore, we propose a multi-prediction guided attention mechanism which provides pixel-wise semantic prediction guidance to better recover the segmentation mask at each scale. This mechanism which transforms the deep supervision to supervised attention is able to guide feature aggregation with more semantic information between intermediate layers. Experiments on the publicly available Duke OCT dataset confirm the effectiveness of the proposed method as well as an improved performance over other state-of-the-art approaches.

eess.IV↗

arXiv2020

Phase Diagram and Structure Map of Binary Nanoparticle Superlattices from a Lennard-Jones Model

A first principle prediction of the binary nanoparticle phase diagram assembled by solvent evaporation has eluded theoretical approaches. In this paper, we show that a binary system interacting through Lennard-Jones (LJ) potential contains all experimental phases in which nanoparticles are effectively described as quasi hard spheres. We report a phase diagram consisting of 53 equilibrium phases, whose stability is quite insensitive to the microscopic details of the potentials, thus giving rise to some type of universality. Furthermore, we show that binary lattices may be understood as consisting of certain particle clusters, i.e. motifs, which provide a generalization of the four conventional Frank-Kasper polyhedral units. Our results show that meta-stable phases share the very same motifs as equilibrium phases. We discuss the connection with packing models, phase diagrams with repulsive potentials and the prediction of likely experimental superlattices.

cond-mat.soft↗

arXiv2020

A Generic Network Compression Framework for Sequential Recommender Systems

Sequential recommender systems (SRS) have become the key technology in capturing user's dynamic interests and generating high-quality recommendations. Current state-of-the-art sequential recommender models are typically based on a sandwich-structured deep neural network, where one or more middle (hidden) layers are placed between the input embedding layer and output softmax layer. In general, these models require a large number of parameters (such as using a large embedding dimension or a deep network architecture) to obtain their optimal performance. Despite the effectiveness, at some point, further increasing model size may be harder for model deployment in resource-constraint devices, resulting in longer responding time and larger memory footprint. To resolve the issues, we propose a compressed sequential recommendation framework, termed as CpRec, where two generic model shrinking techniques are employed. Specifically, we first propose a block-wise adaptive decomposition to approximate the input and softmax matrices by exploiting the fact that items in SRS obey a long-tailed distribution. To reduce the parameters of the middle layers, we introduce three layer-wise parameter sharing schemes. We instantiate CpRec using deep convolutional neural network with dilated kernels given consideration to both recommendation accuracy and efficiency. By the extensive ablation studies, we demonstrate that the proposed CpRec can achieve up to 4$\sim$8 times compression rates in real-world SRS datasets. Meanwhile, CpRec is faster during training\inference, and in most cases outperforms its uncompressed counterpart.

cs.IR↗

arXiv2020

LDA+Usc calculations of phase relations in FeO

Using the LDA+$\textit{U}_\text{sc}$ method, we present calculations phase relations of iron monoxides involving five polytypes in multiple spin-state configurations. The Hubbard parameter $U$ is determined self-consistently simultaneously with the occupation matrix and structures at arbitrary pressures. The Hubbard parameter strongly depends on pressure, structure, and spin state. Comparison with experimental structural data indicates the LDA+$\textit{U}_\text{sc}$ can predict structure, compression curves, phase relations, and transition pressures very well for the insulating B1 and iB8 states. However, it requires additional calculations using the Mermin functional that includes the electronic entropic contribution to the free energy to obtain an nB8 metallic state and a consistent iB8 to nB8 insulator-to-metal transition pressure.

cond-mat.mtrl-sci↗

arXiv2020

The Superconducting Critical Temperature

Two principles govern the critical temperature for superconducting transitions: (1)~intrinsic strength of the pair coupling and (2)~effect of the many-body environment on the efficiency of that coupling. Most discussions take into account only the first but we argue that the properties of unconventional superconductors are governed more often by the second, through dynamical symmetry relating normal and superconducting states. Differentiating these effects is essential to charting a path to the highest-temperature superconductors.

cond-mat.supr-con↗