arXiv Science⌕ Search

arXiv subjects

Yan Liu

Publications and source records attributed to Yan Liu.

At least 487 records · Page 27Linked to original sources
arXiv2016

Latent Space Model for Road Networks to Predict Time-Varying Traffic

Real-time traffic prediction from high-fidelity spatiotemporal traffic sensor datasets is an important problem for intelligent transportation systems and sustainability. However, it is challenging due to the complex topological dependencies and high dynamics associated with changing road conditions. In this paper, we propose a Latent Space Model for Road Networks (LSM-RN) to address these challenges. In particular, given a series of road network snapshots, we learn the attributes of vertices in latent spaces which capture both topological and temporal properties. As these latent attributes are time-dependent, they can estimate how traffic patterns form and evolve. In addition, we present an incremental online algorithm which sequentially and adaptively learn the latent attributes from the temporal graph changes. Our framework enables real-time traffic prediction by 1) exploiting real-time sensor readings to adjust/update the existing latent spaces, and 2) training as data arrives and making predictions on-the-fly with given data. By conducting extensive experiments with a large volume of real-world traffic sensor data, we demonstrate the utility superiority of our framework for real-time traffic prediction on large road networks over competitors as well as a baseline graph-based LSM.

cs.SI↗
arXiv2016

Learning from Multiway Data: Simple and Efficient Tensor Regression

Tensor regression has shown to be advantageous in learning tasks with multi-directional relatedness. Given massive multiway data, traditional methods are often too slow to operate on or suffer from memory bottleneck. In this paper, we introduce subsampled tensor projected gradient to solve the problem. Our algorithm is impressively simple and efficient. It is built upon projected gradient method with fast tensor power iterations, leveraging randomized sketching for further acceleration. Theoretical analysis shows that our algorithm converges to the correct solution in fixed number of iterations. The memory requirement grows linearly with the size of the problem. We demonstrate superior empirical performance on both multi-linear multi-task learning and spatio-temporal applications.

cs.LG↗
arXiv2016

Quantile tests in frequency domain for sinusoid models

For second order stationary processes, the spectral distribution function is uniquely deter- mined by the autocovariance functions of the processes. We define the quantiles of the spectral distribution function and propose two estimators for the quantiles. Asymptotic properties of both estimators are elucidated and the difference from the quantile estimators in time do- main is also indicated. We construct a testing procedure of quantile tests from the asymptotic distribution of the estimators and strong statistical power is shown in our numerical studies.

math.ST↗
arXiv2016

Dipolar matter-wave solitons in two-dimensional anisotropic discrete lattices

We numerically demonstrate two-dimensional (2D) matter-wave solitons in the disk-shaped dipolar Bose-Einstein condensates (BECs) trapped in strongly anisotropic optical lattices (OLs) in a disk's plane. The considered OLs are square lattices which can be formed by interfering two pairs of plane waves with different intensities. The hopping rates of the condensates between two adjacent lattices in the orthogonal directions are different, which gives rise to a linearly anisotropic system. We find that when the polarized orientation of the dipoles is parallel to disk's plane with the same direction, the combined effects of the linearly anisotropy and the nonlocal nonlinear anisotropy strongly influence the formations, as well as the dynamics of the lattice solitons. Particularly, the isotropy-pattern solitons (IPSs) are found when these combined effects reach a balance. Motion, collision and rotation of the IPSs are also studied in detail by means of systematic simulations. We further find that these IPSs can move freely in the 2D anisotropic discrete system, hence giving rise to an anisotropic effective mass. Four types of collisions between the IPSs are identified. By rotating an external magnetic field up to a critical angular velocity, the IPSs can still remain localized and play as a breather. Finally, the influences from the combined effects between the linear and the nonlocal nonlinear anisotropy with consideration of the contact/local nonlinearity are discussed too.

nlin.PS↗
arXiv2016

Information gain versus interference in Bohr's principle of complementarity

Based on modern quantum measurement theory, we use Zurek's "triple model" to study, from the viewpoint of quantum information theory, the wave and particle nature of a photon in a symmetric Mach-Zehnder interferometer. In the process of quantum measurement, the state of both the system and the detector is not an entangled state but a correlated state. We find that the information gain about the photon is related to the correlations (including classical and quantum correlations) between the photon and the detector. We also derive the relationship between the information gain and the fringe visibility. We find that the classical correlations remain consistent with the path distinguishability and can be used to describe the particle-like property of the photon. Quantum correlations are not exactly the same as fringe visibility, but both can represent the quantum coherence of the photon. Finally, we provide an analytical expression for quantum correlations of one type of two-qubit separable states.

quant-ph↗
arXiv2016

Excitation of spin wave modes of a magnetic dot existing in a shifted vortex-state by an out-of-plane oscillating magnetic field

Excitation of spin wave modes of a vortex-state magnetic dot by an out-of-plane oscillating magnetic field is studied numerically in the presence of a static in-plane magnetic field. It is shown, that the application of the in-plane static field shifts the position of the vortex core and leads to the separate excitation of different azimuthal dipolar spin waves my perpendicular oscillating field. It is also shown that the excited dipolar azimuthal spin waves excite the gyrotropic mode of the vortex core rotation and, depending on the excitation frequency, cause a significant modification (increase or decrease) of the apparent dissipation rate of the gyrotropic mode. The last effect can be explained by the nonlinear parametric interaction between the gyrotropic mode and the dipolar spin wave modes with different azimuthal indices.

cond-mat.mes-hall↗
arXiv2016

The DARPA Twitter Bot Challenge

A number of organizations ranging from terrorist groups such as ISIS to politicians and nation states reportedly conduct explicit campaigns to influence opinion on social media, posing a risk to democratic processes. There is thus a growing need to identify and eliminate "influence bots" - realistic, automated identities that illicitly shape discussion on sites like Twitter and Facebook - before they get too influential. Spurred by such events, DARPA held a 4-week competition in February/March 2015 in which multiple teams supported by the DARPA Social Media in Strategic Communications program competed to identify a set of previously identified "influence bots" serving as ground truth on a specific topic within Twitter. Past work regarding influence bots often has difficulty supporting claims about accuracy, since there is limited ground truth (though some exceptions do exist [3,7]). However, with the exception of [3], no past work has looked specifically at identifying influence bots on a specific topic. This paper describes the DARPA Challenge and describes the methods used by the three top-ranked teams.

cs.SI↗
arXiv2016

Quantum phase transition between a topological and a trivial semimetal from holography

We present a holographic model of a topological Weyl semimetal. A key ingredient is a time-reversal breaking parameter and a mass deformation. Upon varying the ratio of mass to time-reversal breaking parameter the model undergoes a quantum phase transition from a topologically nontrivial semimetal to a trivial one. The topological nontrivial semimetal is characterised by the presence of an anomalous Hall effect. The results can be interpreted in terms of the holographic renormalization group (RG) flow leading to restoration of time-reversal at the end point of the RG flow in the trivial phase.

hep-th↗
arXiv2016

Odd viscosity in the quantum critical region of a holographic Weyl semimetal

We study odd viscosity in a holographic model of a Weyl semimetal. The model is characterised by a quantum phase transition from a topological semimetal to a trivial semimetal state. Since the model is axisymmetric in three spatial dimensions there are two independent odd viscosities. Both odd viscosity coefficients are non-vanishing in the quantum critical region and non-zero only due to the mixed axial gravitational anomaly. It is therefore a novel example in which the mixed axial gravitational anomaly gives rise to a transport coefficient at first order in derivatives at finite temperature. We also compute anisotropic shear viscosities and show that one of them violates the KSS bound. In the quantum critical region, the physics of viscosities as well as conductivities is governed by the quantum critical point.

hep-th↗
arXiv2016

A Survey on Social Media Anomaly Detection

Social media anomaly detection is of critical importance to prevent malicious activities such as bullying, terrorist attack planning, and fraud information dissemination. With the recent popularity of social media, new types of anomalous behaviors arise, causing concerns from various parties. While a large amount of work have been dedicated to traditional anomaly detection problems, we observe a surge of research interests in the new realm of social media anomaly detection. In this paper, we present a survey on existing approaches to address this problem. We focus on the new type of anomalous phenomena in the social media and review the recent developed techniques to detect those special types of anomalies. We provide a general overview of the problem domain, common formulations, existing methodologies and potential directions. With this work, we hope to call out the attention from the research community on this challenging problem and open up new directions that we can contribute in the future.

cs.LG↗
arXiv2015

Finite Element Analysis of Electromagnetic Waves in Two-Dimensional Transformed Bianisotropic Media

We analyse wave propagation in two-dimensional bianisotropic media with the Finite Element Method (FEM). We start from the Maxwell-Tellegen's equations in bianisotropic media, and derive some system of coupled Partial Difference Equations (PDEs) for longitudinal electric and magnetic field components. Perfectly Matched Layers (PMLs) are discussed to model such unbounded media. We implement these PDEs and PMLs in a finite element software. We apply transformation optics in order to design some bianisotropic media with interesting functionalities, such as cloaks, concentrators and rotators. We propose a design of metamaterial with concentric layers made of homogeneous media with isotropic permittivity, permeability and magneto-electric parameters that mimic the required effective anisotropic tensors of a bianisotropic cloak in the long wavelength limit (homogenization approach). Our numerical results show that well-known metamaterials can be transposed to bianisotropic media.

physics.optics↗
arXiv2015

Distilling Knowledge from Deep Networks with Applications to Healthcare Domain

Exponential growth in Electronic Healthcare Records (EHR) has resulted in new opportunities and urgent needs for discovery of meaningful data-driven representations and patterns of diseases in Computational Phenotyping research. Deep Learning models have shown superior performance for robust prediction in computational phenotyping tasks, but suffer from the issue of model interpretability which is crucial for clinicians involved in decision-making. In this paper, we introduce a novel knowledge-distillation approach called Interpretable Mimic Learning, to learn interpretable phenotype features for making robust prediction while mimicking the performance of deep learning models. Our framework uses Gradient Boosting Trees to learn interpretable features from deep learning models such as Stacked Denoising Autoencoder and Long Short-Term Memory. Exhaustive experiments on a real-world clinical time-series dataset show that our method obtains similar or better performance than the deep learning models, and it provides interpretable phenotypes for clinical decision making.

stat.ML↗
arXiv2015

Weighted Schatten $p$-Norm Minimization for Image Denoising and Background Subtraction

Low rank matrix approximation (LRMA), which aims to recover the underlying low rank matrix from its degraded observation, has a wide range of applications in computer vision. The latest LRMA methods resort to using the nuclear norm minimization (NNM) as a convex relaxation of the nonconvex rank minimization. However, NNM tends to over-shrink the rank components and treats the different rank components equally, limiting its flexibility in practical applications. We propose a more flexible model, namely the Weighted Schatten $p$-Norm Minimization (WSNM), to generalize the NNM to the Schatten $p$-norm minimization with weights assigned to different singular values. The proposed WSNM not only gives better approximation to the original low-rank assumption, but also considers the importance of different rank components. We analyze the solution of WSNM and prove that, under certain weights permutation, WSNM can be equivalently transformed into independent non-convex $l_p$-norm subproblems, whose global optimum can be efficiently solved by generalized iterated shrinkage algorithm. We apply WSNM to typical low-level vision problems, e.g., image denoising and background subtraction. Extensive experimental results show, both qualitatively and quantitatively, that the proposed WSNM can more effectively remove noise, and model complex and dynamic scenes compared with state-of-the-art methods.

cs.CV↗
arXiv2015

An empirical likelihood approach for symmetric $α$-stable processes

Empirical likelihood approach is one of non-parametric statistical methods, which is applied to the hypothesis testing or construction of confidence regions for pivotal unknown quantities. This method has been applied to the case of independent identically distributed random variables and second order stationary processes. In recent years, we observe heavy-tailed data in many fields. To model such data suitably, we consider symmetric scalar and multivariate $α$-stable linear processes generated by infinite variance innovation sequence. We use a Whittle likelihood type estimating function in the empirical likelihood ratio function and derive the asymptotic distribution of the empirical likelihood ratio statistic for $α$-stable linear processes. With the empirical likelihood statistic approach, the theory of estimation and testing for second order stationary processes is nicely extended to heavy-tailed data analyses, not straightforward, and applicable to a lot of financial statistical analyses.

math.ST↗
arXiv2015

Quantum correlations of two qubits interacting with a macroscopic medium

We consider two particles of spin-1/2 interacting with a one-dimensional N-spin array, which is an exactly solvable model. The dynamics of entanglement and quantum discord (QD) of the spins of the two particles is investigated by regarding the 1D N-spin array as the environment. It is found that although the entanglement may suffers a sudden death and a sudden birth in the evolution, it can neither be generated nor become larger than its initial value. Different from the entanglement dynamics, QD can be amplified, and even be generated by the interaction between particles and the common environment. We also observe that QD decays asymptotically to zero and later experiences a rival when the average number of excitation in the 1D N-spin array becomes larger in the case of nonzero inter-distance between two particles.

quant-ph↗
arXiv2015

Fast binarized time-reversed adapted-perturbation (b-TRAP) optical focusing inside scattering media

Light scattering inhibits high-resolution optical imaging, manipulation and therapy deep inside biological tissue by preventing focusing. To form deep foci, wavefront-shaping and time-reversal techniques that break the optical diffusion limit have been developed. For in vivo applications, such focusing must provide high gain, high speed, and a large number of spatial modes. However, none of the previous techniques meet these requirements simultaneously. Here, we overcome this challenge by rapidly measuring the perturbed optical field within a single camera exposure followed by adaptively time-reversing the phase-binarized perturbation. Consequently, a phase-conjugated wavefront is synthesized within a millisecond, two orders of magnitude shorter than the digitally achieved record. We demonstrated real-time focusing in dynamic scattering media, and extended laser speckle contrast imaging to new depths. The unprecedented combination of fast response, high gain, and large mode count makes this work a major stride toward in vivo deep tissue optical imaging, manipulation, and therapy.

physics.optics↗