Search arXiv⌕ Search

arXiv subjects

Li Chen

Publications and source records attributed to Li Chen.

At least 631 records · Page 35Linked to original sources
arXiv2018

Programmable Switch as a Parallel Computing Device

Modern switches have packet processing capacity of up to multi-tera bits per second, and they are also becoming more and more programmable. We seek to understand whether the programmability can translate packet processing capacity to computational power for parallel computing applications. In this paper, we first develop a simple mathematical model to understand the costs and overheads of data plane computation. Then we validate the the performance benefits of offloading computation to network. Using experiments on real data center network, we finnd that offloading computation to the data plane results in up to 20x speed-up for a simple Map-Reduce application. Motivated by this, we propose a parallel programming framework, p4mr, to help users efficiently program multiple switches. We successfully build and test a prototype of p4mr on a simulated testbed.

cs.NI↗
arXiv2018

Shield: Fast, Practical Defense and Vaccination for Deep Learning using JPEG Compression

The rapidly growing body of research in adversarial machine learning has demonstrated that deep neural networks (DNNs) are highly vulnerable to adversarially generated images. This underscores the urgent need for practical defense that can be readily deployed to combat attacks in real-time. Observing that many attack strategies aim to perturb image pixels in ways that are visually imperceptible, we place JPEG compression at the core of our proposed Shield defense framework, utilizing its capability to effectively "compress away" such pixel manipulation. To immunize a DNN model from artifacts introduced by compression, Shield "vaccinates" a model by re-training it with compressed images, where different compression levels are applied to generate multiple vaccinated models that are ultimately used together in an ensemble defense. On top of that, Shield adds an additional layer of protection by employing randomization at test time that compresses different regions of an image using random compression levels, making it harder for an adversary to estimate the transformation performed. This novel combination of vaccination, ensembling, and randomization makes Shield a fortified multi-pronged protection. We conducted extensive, large-scale experiments using the ImageNet dataset, and show that our approaches eliminate up to 94% of black-box attacks and 98% of gray-box attacks delivered by the recent, strongest attacks, such as Carlini-Wagner's L2 and DeepFool. Our approaches are fast and work without requiring knowledge about the model.

cs.CV↗
arXiv2018

Gundy-Varopoulos martingale transforms and their projection operators on manifolds and vector bundles

This paper proves the $L^p$ boundedness of generalized first order Riesz transforms obtained as conditional expectations of martingale transforms à la Gundy-Varopoulos for quite general diffusions on manifolds and vector bundles. Several specific examples and applications are presented: Lie groups of compact type, the Heisenberg group, SU(2), and Riesz transforms on forms and spinors.

math.FA↗
arXiv2018

Nonlocal nonlinear reaction preventing blow-up in Keller-Segel system

This paper is devoted to the analysis of non-negative solutions for the chemotaxis model with nonlocal source in bounded domain. The qualitative behavior of solutions is determined by the nonlinearity from the aggregation and the reaction. The nonlocal nonlinear Fisher-KPP reaction helps preventing blow-up phenomena in chemotaxis system. For appropriately chosen exponents, the global existence of classical solutions is proved with arbitrary initial data, where a modified Moser-Alikakos iteration method plays a key role in the a priori estimates.

math.AP↗
arXiv2018

Rate of Convergence towards Hartree Dynamics with Singular Interaction Potential

We consider a system of $N$-Bosons with a two-body interaction potential $V \in L^2(\mathbb{R}^3)+L^\infty(\mathbb{R}^3)$, possibly singular than the Coulomb interaction. We show that, with $H^1(\mathbb{R}^3)$ initial data, the difference between the many-body Schrödinger evolution in the mean-field regime and the corresponding Hartree dynamics is of order $1/N$, for any fixed time. The $N$-dependence of the bound is optimal.

math-ph↗
arXiv2018

Modeling of a diffusion with aggregation: rigorous derivation and numerical simulation

In this paper, a diffusion-aggregation equation with delta potential is introduced. Based on the global existence and uniform estimates of solutions to the diffusion-aggregation equation, we also provide the rigorous derivation from a stochastic particle system while introducing an intermediate particle system with smooth interaction potential. The theoretical results are compared to numerical simulations relying on suitable discretization schemes for the microscopic and macroscopic level. In particular, the regime switch where the analytic theory fails is numerically analyzed very carefully and allows for a better understanding of the equation.

math.AP↗
arXiv2018

Asteroseismology of KIC 7107778: a binary comprising almost identical subgiants

We analyze an asteroseismic binary system: KIC 7107778, a non-eclipsing, unresolved target, with solar-like oscillations in both components. We used \emph{Kepler} short cadence time series spanning nearly two years to obtain the power spectrum. Oscillation mode parameters were determined using Bayesian inference and a nested sampling Monte Carlo algorithm with the \texttt{DIAMONDS} package. The power profiles of the two components fully overlap, indicating their close similarity. We modelled the two stars with \texttt{MESA} and calculated oscillation frequencies with \texttt{GYRE}. Stellar fundamental parameters (mass, radius and age) were estimated by grid modelling with atmospheric parameters and the oscillation frequencies of $l=0,2$ modes as constraints. Most $l=1$ mixed modes were identified with models searched using a bisection method. Stellar parameters for the two sub-giant stars are: $M_A=1.42\pm0.06$ $M_{\astrosun}$, $M_B=1.39\pm0.03$ $M_{\astrosun}$, $R_A=2.93\pm0.05$ $R_{\astrosun}$, $R_B=2.76\pm0.04$ $R_{\astrosun}$, $t_A=3.32\pm0.54$ Gyr and $t_B=3.51\pm0.33$ Gyr. The mass difference of the system is $\sim$1\%. The results confirm their simultaneous birth and evolution, as is expected from binary formation. KIC 7107778 comprises almost identical twins, and is the first asteroseismic sub-giant binary to be detected.

astro-ph.SR↗
arXiv2018

Supercritical degenerate parabolic-parabolic Keller-Segel system -- existence criterion given by the best constant in Sobolev's inequality

This article presents a relationship between the sharp constant of the Sobolev inequality and the initial criterion to the global existence of degenerate parabolic-parabolic Keller-Segel system with the diffusion exponent $\frac{2n}{2+n}<m<2-\frac{2}{n}$. The global weak solution obtained in this article does not need any smallness assumption on the initial data. Furthermore, a uniform in time $L^{\infty}$ estimate of the weak solutions is obtained via the Moser iteration, where the constant in $L^p$ estimate for the gradient of the chemical concentration has been exactly formulated in order to complete the iteration process.

math.AP↗
arXiv2018

HeNet: A Deep Learning Approach on Intel$^\circledR$ Processor Trace for Effective Exploit Detection

This paper presents HeNet, a hierarchical ensemble neural network, applied to classify hardware-generated control flow traces for malware detection. Deep learning-based malware detection has so far focused on analyzing executable files and runtime API calls. Static code analysis approaches face challenges due to obfuscated code and adversarial perturbations. Behavioral data collected during execution is more difficult to obfuscate but recent research has shown successful attacks against API call based malware classifiers. We investigate control flow based characterization of a program execution to build robust deep learning malware classifiers. HeNet consists of a low-level behavior model and a top-level ensemble model. The low-level model is a per-application behavior model, trained via transfer learning on a time-series of images generated from control flow trace of an execution. We use Intel$^\circledR$ Processor Trace enabled processor for low overhead execution tracing and design a lightweight image conversion and segmentation of the control flow trace. The top-level ensemble model aggregates the behavior classification of all the trace segments and detects an attack. The use of hardware trace adds portability to our system and the use of deep learning eliminates the manual effort of feature engineering. We evaluate HeNet against real-world exploitations of PDF readers. HeNet achieves 100\% accuracy and 0\% false positive on test set, and higher classification accuracy compared to classical machine learning algorithms.

cs.CR↗
arXiv2017

New Open Cluster Candidates Discovered in the XSTPS-GAC Survey

The Xuyi Schmidt Telescope Photometric Survey of the Galactic Anti-center (XSTPS-GAC) is a photometric sky survey that covers nearly 6 000 deg^2 towards Galactic anti-center in g r i bands. Half of its survey field locates on the Galactic Anti-center disk, which makes XSTPS-GAC highly suitable for searching new open clusters in the GAC region. In this paper, we report new open cluster candidates discovered in this survey, as well as properties of these open cluster candidates, such as age, distance and reddening, derived by isochrone fitting in the color-magnitude diagram (CMD). These open cluster candidates are stellar density peaks detected in the star density maps by applying the method from Koposov et al. (2008). Each candidate is inspected on its true color image composed from XSTPS-GAC three band images. Then its CMD is checked, in order to identify whether the central region stars have a clear isochrone-like trend differing from the background stars. The parameters derived from isochrone fitting for these candidates are mainly based on three band photometry of XSTPS-GAC. Meanwhile, when these new candidates are able to be seen clearly on 2MASS, their parameters are also derived based on 2MASS (J-H, J) CMD. Finally, there are 320 known open clusters rediscovered and 24 new open cluster candidates discovered in this work. Further more, the parameters of these new candidates, as well as another 11 known recovered open clusters, are properly determined for the first time.

astro-ph.SR↗
arXiv2017

Y-net: 3D intracranial artery segmentation using a convolutional autoencoder

Automated segmentation of intracranial arteries on magnetic resonance angiography (MRA) allows for quantification of cerebrovascular features, which provides tools for understanding aging and pathophysiological adaptations of the cerebrovascular system. Using a convolutional autoencoder (CAE) for segmentation is promising as it takes advantage of the autoencoder structure in effective noise reduction and feature extraction by representing high dimensional information with low dimensional latent variables. In this report, an optimized CAE model (Y-net) was trained to learn a 3D segmentation model of intracranial arteries from 49 cases of MRA data. The trained model was shown to perform better than the three traditional segmentation methods in both binary classification and visual evaluation.

eess.IV↗
arXiv2017

Chemotaxis model with subcritical exponent in nonlocal reaction

This paper deals with a parabolic-elliptic chemotaxis system with nonlocal type of source in the whole space. It's proved that the initial value problem possesses a unique global solution which is uniformly bounded. Here we identify the exponents regimes of nonlinear reaction and aggregation in such a way that their scaling and the diffusion term coincide (see Introduction). Comparing to the classical KS model (without the source term), it's shown that how energy estimates give natural conditions on the nonlinearities implying the absence of blow-up for the solution without any restriction on the initial data.

math.AP↗
arXiv2017

Effect of Notch Structure on Magnetic Domain Movement in Planar Nanowires

We present the direct observation of magnetic domain motion in permalloy nanowires with notches using a wide-field Kerr microscopy technique. The domain wall motion can be modulated by the size and shape of the notch structure in the nanowires. It is demonstrated that the coercive fields can be tuned by modulating the notches. The experimental results are consistent with the micro-magnetic simulation results. The relationship between the notch angle and the domain nuclease are also studied. This work is useful for the design and development of the notch-based spintronic devices.

cond-mat.mes-hall↗
arXiv2017

Insight-HXMT observations of the first binary neutron star merger GW170817

Finding the electromagnetic (EM) counterpart of binary compact star merger, especially the binary neutron star (BNS) merger, is critically important for gravitational wave (GW) astronomy, cosmology and fundamental physics. On Aug. 17, 2017, Advanced LIGO and \textit{Fermi}/GBM independently triggered the first BNS merger, GW170817, and its high energy EM counterpart, GRB 170817A, respectively, resulting in a global observation campaign covering gamma-ray, X-ray, UV, optical, IR, radio as well as neutrinos. The High Energy X-ray telescope (HE) onboard \textit{Insight}-HXMT (Hard X-ray Modulation Telescope) is the unique high-energy gamma-ray telescope that monitored the entire GW localization area and especially the optical counterpart (SSS17a/AT2017gfo) with very large collection area ($\sim$1000 cm$^2$) and microsecond time resolution in 0.2-5 MeV. In addition, \textit{Insight}-HXMT quickly implemented a Target of Opportunity (ToO) observation to scan the GW localization area for potential X-ray emission from the GW source. Although it did not detect any significant high energy (0.2-5 MeV) radiation from GW170817, its observation helped to confirm the unexpected weak and soft nature of GRB 170817A. Meanwhile, \textit{Insight}-HXMT/HE provides one of the most stringent constraints (~10$^{-7}$ to 10$^{-6}$ erg/cm$^2$/s) for both GRB170817A and any other possible precursor or extended emissions in 0.2-5 MeV, which help us to better understand the properties of EM radiation from this BNS merger. Therefore the observation of \textit{Insight}-HXMT constitutes an important chapter in the full context of multi-wavelength and multi-messenger observation of this historical GW event.

astro-ph.HE↗
arXiv2017

Quantum Monte Carlo study of the metal to insulator transition on a honeycomb lattice with 1/r interactions

Describing correlated electron systems near phase transitions has been a major challenge in computational condensed-matter physics. In this paper, we apply highly accurate fixed node quantum Monte Carlo techniques, which directly work with many body wave functions and simulate electron correlations, to investigate the metal to insulator transition of a correlated hydrogen lattice. By calculating spin and charge properties, and analyzing the low energy Hilbert space, we identify the transition point and identify order parameters that can be used to detect the transition. Our results provide a benchmark for density functional theories seeking to treat correlated electron systems.

cond-mat.str-el↗
arXiv2017

Riesz transforms for bounded Laplacians on graphs

We study several problems related to the $\ell^p$ boundedness of Riesz transforms for graphs endowed with so-called bounded Laplacians. Introducing a proper notion of gradient of functions on edges, we prove for $p\in(1,2]$ an $\ell^p$ estimate for the gradient of the continuous time heat semigroup, an $\ell^p$ interpolation inequality as well as the $\ell^p$ boundedness of the modified Littlewood-Paley-Stein functions for all graphs with bounded Laplacians. This yields an analogue to Dungey's results in [Dungey08] while removing some additional assumptions. Coming back to the classical notion of gradient, we give a counterexample to the interpolation inequality hence to the boundedness of Riesz transforms for bounded Laplacians for $1<p<2$. Finally, we prove the boundedness of the Riesz transform for $1< p<\infty$ under the assumption of positive spectral gap.

math.MG↗
arXiv2017

The Evolution of the Phase Lags Associated with the Type-C Quasi-periodic Oscillation in GX 339--4 during the 2006/2007 Outburst

We present the evolution of the phase lags associated with the type-C QPO in GX 339--4 during the rising phase of the 2006/2007 outburst. We find that the phase lags at the QPO frequency are always positive (hard), and show very different behavior between QPOs with frequencies below and above $\sim1.7$ Hz: when the QPO frequency is below $\sim1.7$ Hz, the phase lags increase both with QPO frequency and energy, while when the QPO frequency is above $\sim1.7$ Hz, the phase lags remain more or less constant. When the QPO frequency is higher than $\sim1.7$ Hz, a broad feature is always present in the lag-energy spectra at around 6.5 keV, suggesting that the reflection component may have a significant contribution to the phase lags. Below $\sim1.7$ Hz, the QPO rms first decreases with energy and then turns to almost flat, while above $\sim1.7$ Hz, the QPO rms increases with energy. During the transition from the low-hard state to the hard-intermediate state, the second harmonic and subharmonic of this QPO appear in the power density spectra. The second-harmonic and subharmonic phase lags show very similar evolution with their centroid frequencies. However, the energy dependence of the second-harmonic and subharmonic phase lags are quite different. Our results suggest that, at different phases of the outburst, different mechanisms may be responsible for the phase lags of the QPO. We briefly discuss the possible scenarios for producing the lags.

astro-ph.HE↗