arXiv subjects

Cheng Chen

Publications and source records attributed to Cheng Chen.

At least 343 records · Page 19Linked to original sources

arXiv2020

2D-Material-Assisted Bistable Switching of Gap Plasmons Disclosed By Femtosecond Pulse Scattering Spectra

Nanosphere-on-mirror plasmonic antennas, each having a monolayer graphene or MoS2 sheet in the gap, were pumped with a femtosecond laser. Abrupt turnings in the scattering linewidth and peak intensity trends as the laser power changed were experimentally observed. Theoretical modelling of dynamic plasmon evolvement attributes the turning to transitioning between two plasmon states, with a universal switching threshold of four-wave mixing efficiency around 0.14%. This bistability is rendered by a strong feedback from the nonlinear gap current to the gap plasmons, and involvement of both four and high-order wave mixing. This work reveals a pathway to making energy efficient nonlinear plasmonic elements.

physics.optics↗

arXiv2020

Revisiting Co-Occurring Directions: Sharper Analysis and Efficient Algorithm for Sparse Matrices

We study the streaming model for approximate matrix multiplication (AMM). We are interested in the scenario that the algorithm can only take one pass over the data with limited memory. The state-of-the-art deterministic sketching algorithm for streaming AMM is the co-occurring directions (COD), which has much smaller approximation errors than randomized algorithms and outperforms other deterministic sketching methods empirically. In this paper, we provide a tighter error bound for COD whose leading term considers the potential approximate low-rank structure and the correlation of input matrices. We prove COD is space optimal with respect to our improved error bound. We also propose a variant of COD for sparse matrices with theoretical guarantees. The experiments on real-world sparse datasets show that the proposed algorithm is more efficient than baseline methods.

cs.LG↗

arXiv2020

FewJoint: A Few-shot Learning Benchmark for Joint Language Understanding

Few-shot learning (FSL) is one of the key future steps in machine learning and has raised a lot of attention. However, in contrast to the rapid development in other domains, such as Computer Vision, the progress of FSL in Nature Language Processing (NLP) is much slower. One of the key reasons for this is the lacking of public benchmarks. NLP FSL researches always report new results on their own constructed few-shot datasets, which is pretty inefficient in results comparison and thus impedes cumulative progress. In this paper, we present FewJoint, a novel Few-Shot Learning benchmark for NLP. Different from most NLP FSL research that only focus on simple N-classification problems, our benchmark introduces few-shot joint dialogue language understanding, which additionally covers the structure prediction and multi-task reliance problems. This allows our benchmark to reflect the real-word NLP complexity beyond simple N-classification. Our benchmark is used in the few-shot learning contest of SMP2020-ECDT task-1. We also provide a compatible FSL platform to ease experiment set-up.

cs.CL↗

arXiv2020

C2C-GenDA: Cluster-to-Cluster Generation for Data Augmentation of Slot Filling

Slot filling, a fundamental module of spoken language understanding, often suffers from insufficient quantity and diversity of training data. To remedy this, we propose a novel Cluster-to-Cluster generation framework for Data Augmentation (DA), named C2C-GenDA. It enlarges the training set by reconstructing existing utterances into alternative expressions while keeping semantic. Different from previous DA works that reconstruct utterances one by one independently, C2C-GenDA jointly encodes multiple existing utterances of the same semantics and simultaneously decodes multiple unseen expressions. Jointly generating multiple new utterances allows to consider the relations between generated instances and encourages diversity. Besides, encoding multiple existing utterances endows C2C with a wider view of existing expressions, helping to reduce generation that duplicates existing data. Experiments on ATIS and Snips datasets show that instances augmented by C2C-GenDA improve slot filling by 7.99 (11.9%) and 5.76 (13.6%) F-scores respectively, when there are only hundreds of training utterances.

cs.CL↗

arXiv2020

Neural Rate Control for Video Encoding using Imitation Learning

In modern video encoders, rate control is a critical component and has been heavily engineered. It decides how many bits to spend to encode each frame, in order to optimize the rate-distortion trade-off over all video frames. This is a challenging constrained planning problem because of the complex dependency among decisions for different video frames and the bitrate constraint defined at the end of the episode. We formulate the rate control problem as a Partially Observable Markov Decision Process (POMDP), and apply imitation learning to learn a neural rate control policy. We demonstrate that by learning from optimal video encoding trajectories obtained through evolution strategies, our learned policy achieves better encoding efficiency and has minimal constraint violation. In addition to imitating the optimal actions, we find that additional auxiliary losses, data augmentation/refinement and inference-time policy improvements are critical for learning a good rate control policy. We evaluate the learned policy against the rate control policy in libvpx, a widely adopted open source VP9 codec library, in the two-pass variable bitrate (VBR) mode. We show that over a diverse set of real-world videos, our learned policy achieves 8.5% median bitrate reduction without sacrificing video quality.

cs.LG↗

arXiv2020

Neural Network Training Techniques Regularize Optimization Trajectory: An Empirical Study

Modern deep neural network (DNN) trainings utilize various training techniques, e.g., nonlinear activation functions, batch normalization, skip-connections, etc. Despite their effectiveness, it is still mysterious how they help accelerate DNN trainings in practice. In this paper, we provide an empirical study of the regularization effect of these training techniques on DNN optimization. Specifically, we find that the optimization trajectories of successful DNN trainings consistently obey a certain regularity principle that regularizes the model update direction to be aligned with the trajectory direction. Theoretically, we show that such a regularity principle leads to a convergence guarantee in nonconvex optimization and the convergence rate depends on a regularization parameter. Empirically, we find that DNN trainings that apply the training techniques achieve a fast convergence and obey the regularity principle with a large regularization parameter, implying that the model updates are well aligned with the trajectory. On the other hand, DNN trainings without the training techniques have slow convergence and obey the regularity principle with a small regularization parameter, implying that the model updates are not well aligned with the trajectory. Therefore, different training techniques regularize the model update direction via the regularity principle to facilitate the convergence.

cs.LG↗

arXiv2020

The evolution of a circumplanetary disc with a dead zone

We investigate whether the regular Galilean satellites could have formed in the dead zone of a circumplanetary disc. A dead zone is a region of weak turbulence in which the magnetorotational instability (MRI) is suppressed, potentially an ideal environment for satellite formation. With the grid-based hydrodynamic code, FARGO3D, we examine the evolution of a circumplanetary disc model with a dead zone. Material accumulates in the dead zone of the disc leading to a higher total mass and but a similar temperature profile compared to a fully turbulent disc model. The tidal torque increases the rate of mass transport through the dead zone leading to a steady state disc with a dead zone that does not undergo accretion outbursts. We explore a range of disc, dead zone and mass inflow parameters and find that the maximum mass of the disc is around 0.001 MJ . Since the total solid mass of such a disc is much lower, we find that there is not sufficient material in the disc for in situ formation of the Galilean satellites and that external supplement is required.

astro-ph.EP↗

arXiv2020

Efficient Projection-Free Algorithms for Saddle Point Problems

The Frank-Wolfe algorithm is a classic method for constrained optimization problems. It has recently been popular in many machine learning applications because its projection-free property leads to more efficient iterations. In this paper, we study projection-free algorithms for convex-strongly-concave saddle point problems with complicated constraints. Our method combines Conditional Gradient Sliding with Mirror-Prox and shows that it only requires $\tilde{O}(1/\sqrtε)$ gradient evaluations and $\tilde{O}(1/ε^2)$ linear optimizations in the batch setting. We also extend our method to the stochastic setting and propose first stochastic projection-free algorithms for saddle point problems. Experimental results demonstrate the effectiveness of our algorithms and verify our theoretical guarantees.

math.OC↗

arXiv2020

Anomalous linear magnetoresistance in high quality crystalline lead thin films

Intriguing novel phenomena in lead films inspire new understanding of quantum physics, such as quantum size effect and quantum phase transitions etc. The improvement of the sample quality makes it even more promising to explore the intrinsic properties in two-dimensional system. In this paper, we show that the crystalline interfacial striped incommensurate layer can increase the quality of the lead films and significantly enhance the magnitude of magnetoresistance. By performing systematic transport measurement, a predominant anomalous linear magnetoresistance is revealed, and the widely used Parish-Littlewood model and Abrikosov's explanation fail to describe the observation. Instead, we propose a new model of linear magnetoresistance based on linear band structure, which shows a good agreement with the experimental results. Our studies reveal a novel origin of linear magnetoresistance which may also be helpful to understand the linear magnetoresistance in other materials with linear dispersion of electronic structure.

cond-mat.mes-hall↗

arXiv2020

Data Augmentation at the LHC through Analysis-specific Fast Simulation with Deep Learning

We present a fast simulation application based on a Deep Neural Network, designed to create large analysis-specific datasets. Taking as an example the generation of W+jet events produced in sqrt(s)= 13 TeV proton-proton collisions, we train a neural network to model detector resolution effects as a transfer function acting on an analysis-specific set of relevant features, computed at generation level, i.e., in absence of detector effects. Based on this model, we propose a novel fast-simulation workflow that starts from a large amount of generator-level events to deliver large analysis-specific samples. The adoption of this approach would result in about an order-of-magnitude reduction in computing and storage requirements for the collision simulation workflow. This strategy could help the high energy physics community to face the computing challenges of the future High-Luminosity LHC.

physics.comp-ph↗

arXiv2020

FedCluster: Boosting the Convergence of Federated Learning via Cluster-Cycling

We develop FedCluster--a novel federated learning framework with improved optimization efficiency, and investigate its theoretical convergence properties. The FedCluster groups the devices into multiple clusters that perform federated learning cyclically in each learning round. Therefore, each learning round of FedCluster consists of multiple cycles of meta-update that boost the overall convergence. In nonconvex optimization, we show that FedCluster with the devices implementing the local {stochastic gradient descent (SGD)} algorithm achieves a faster convergence rate than the conventional {federated averaging (FedAvg)} algorithm in the presence of device-level data heterogeneity. We conduct experiments on deep learning applications and demonstrate that FedCluster converges significantly faster than the conventional federated learning under diverse levels of device-level data heterogeneity for a variety of local optimizers.

cs.LG↗

arXiv2020

Functional Linear Regression: Dependence and Error Contamination

Functional linear regression is an important topic in functional data analysis. It is commonly assumed that samples of the functional predictor are independent realizations of an underlying stochastic process, and are observed over a grid of points contaminated by i.i.d. measurement errors. In practice, however, the dynamical dependence across different curves may exist and the parametric assumption on the error covariance structure could be unrealistic. In this paper, we consider functional linear regression with serially dependent observations of the functional predictor, when the contamination of the predictor by the white noise is genuinely functional with fully nonparametric covariance structure. Inspired by the fact that the autocovariance function of observed functional predictors automatically filters out the impact from the unobservable noise term, we propose a novel autocovariance-based generalized method-of-moments estimate of the slope function. We also develop a nonparametric smoothing approach to handle the scenario of partially observed functional predictors. The asymptotic properties of the resulting estimators under different scenarios are established. Finally, we demonstrate that our proposed method significantly outperforms possible competing methods through an extensive set of simulations and an analysis of a public financial dataset.

stat.ME↗

arXiv2020

Observation of topological electronic structure in quasi-1D superconductor TaSe3

Topological superconductors (TSCs), with the capability to host Majorana bound states that can lead to non-Abelian statistics and application in quantum computation, have been one of the most intensively studied topics in condensed matter physics recently. Up to date, only a few compounds have been proposed as candidates of intrinsic TSCs, such as doped topological insulator CuxBi2Se3 and iron-based superconductor FeTe0.55Se0.45. Here, by carrying out synchrotron and laser based angle-resolved photoemission spectroscopy (ARPES), we systematically investigated the electronic structure of a quasi-1D superconductor TaSe3, and identified the nontrivial topological surface states. In addition, our scanning tunneling microscopy (STM) study revealed a clean cleaved surface with a persistent superconducting gap, proving it suitable for further investigation of potential Majorana modes. These results prove TaSe3 as a stoichiometric TSC candidate that is stable and exfoliable, therefore a great platform for the study of rich novel phenomena and application potentials.

cond-mat.mtrl-sci↗

arXiv2020

Learning Directional Feature Maps for Cardiac MRI Segmentation

Cardiac MRI segmentation plays a crucial role in clinical diagnosis for evaluating personalized cardiac performance parameters. Due to the indistinct boundaries and heterogeneous intensity distributions in the cardiac MRI, most existing methods still suffer from two aspects of challenges: inter-class indistinction and intra-class inconsistency. To tackle these two problems, we propose a novel method to exploit the directional feature maps, which can simultaneously strengthen the differences between classes and the similarities within classes. Specifically, we perform cardiac segmentation and learn a direction field pointing away from the nearest cardiac tissue boundary to each pixel via a direction field (DF) module. Based on the learned direction field, we then propose a feature rectification and fusion (FRF) module to improve the original segmentation features, and obtain the final segmentation. The proposed modules are simple yet effective and can be flexibly added to any existing segmentation network without excessively increasing time and space complexity. We evaluate the proposed method on the 2017 MICCAI Automated Cardiac Diagnosis Challenge (ACDC) dataset and a large-scale self-collected dataset, showing good segmentation performance and robust generalization ability of the proposed method.

cs.CV↗

arXiv2020

High-resolution imaging of Rydberg atoms in optical lattices using an aspheric-lens objective in vacuum

We present a high-resolution, simple and versatile system for imaging ultracold Rydberg atoms in optical lattices. The imaging objective is a single aspheric lens (with a working distance of 20.6 mm and a numerical aperture (NA) of 0.51) placed inside the vacuum chamber. Adopting a large-working-distance lens leaves room for electrodes and electrostatic shields to control electric fields around Rydberg atoms. With this setup, we achieve an Rayleigh resolution of 1.10 $μ$m or $1.41λ$ ($λ=780$ nm), limited by the NA of the aspheric lens. For systems of highly excited Rydberg states with blockade radii greater than a few $μ$m, the resolution achieved is sufficient for studying many physical processes of interest.

cond-mat.quant-gas↗

arXiv2020

Electronic structure of a Si-containing topological Dirac semimetal CaAl2Si2

There has been an upsurge in the discovery of topological quantum materials, where various topological insulators and semimetals have been theoretically predicted and experimentally observed. However, only very few of them contains silicon, the most widely used element in electronic industry. Recently, ternary compound CaAl2Si2 has been predicted to be a topological Dirac semimetal, hosting Lorentz-symmetry-violating quasiparticles with a strongly tilted conical band dispersion. In this work, by using high-resolution angle-resolved photoemission spectroscopy (ARPES), we investigated the comprehensive electronic structure of CaAl2Si2. A pair of topological Dirac crossings is observed along the kz direction, in good agreement with the ab initio calculations, confirming the topological Dirac semimetal nature of the compound. Our study expands the topological material family on Si-containing compounds, which have great application potential in realizing low-cost, nontoxic electronic device with topological quantum states.

cond-mat.mtrl-sci↗

arXiv2020

Study of single-particle resonant states with Green's function method

The relativistic mean field theory with the Green's function method is taken to study the single-particle resonant states. Different from our previous work [Phys.Rev.C 90,054321(2014)], the resonant states are identified by searching for the poles of Green's function or the extremes of the density of states. This new approach is very effective for all kinds of resonant states, no matter it is broad or narrow. The dependence on the space size for the resonant energies, widths, and the density distributions in the coordinate space has been checked and it is found very stable. Taking $^{120}$Sn as an example, four new broad resonant states $2g_{7/2}$, $2g_{9/2}$, $2h_{11/2}$ and $1j_{13/2}$ are observed, and also the accuracy for the width of the very narrow resonant state $1h_{9/2}$ is highly improved to be $1\times 10^{-8}$ MeV. Besides, our results are very close to those by the complex momentum representation method and the complex scaling method.

nucl-th↗

arXiv2020

Polar planets around highly eccentric binaries are the most stable

We study the orbital stability of a non-zero mass, close-in circular orbit planet around an eccentric orbit binary for various initial values of the binary eccentricity, binary mass fraction, planet mass, planet semi--major axis, and planet inclination by means of numerical simulations that cover $5 \times 10^4$ binary orbits. For small binary eccentricity, the stable orbits that extend closest to the binary (most stable orbits) are nearly retrograde and circulating. For high binary eccentricity, the most stable orbits are highly inclined and librate near the so-called generalised polar orbit which is a stationary orbit that is fixed in the frame of the binary orbit. For more extreme mass ratio binaries, there is a greater variation in the size of the stability region (defined by initial orbital radius and inclination) with planet mass and initial inclination, especially for low binary eccentricity. For low binary eccentricity, inclined planet orbits may be unstable even at large orbital radii (separation $> 5 \,a_{\rm b}$). The escape time for an unstable planet is generally shorter around an equal mass binary compared with an unequal mass binary. Our results have implications for circumbinary planet formation and evolution and will be helpful for understanding future circumbinary planet observations.

astro-ph.EP↗