Z. -T. Lu

Z. -T. Lu
Are you Z. -T. Lu?

Claim your profile, edit publications, add additional information:

Contact Details

Name
Z. -T. Lu
Affiliation
Location

Pubs By Year

Pub Categories

 
Computer Science - Computation and Language (11)
 
Physics - Materials Science (7)
 
High Energy Physics - Phenomenology (6)
 
Computer Science - Learning (6)
 
High Energy Physics - Experiment (4)
 
Computer Science - Artificial Intelligence (3)
 
Statistics - Machine Learning (3)
 
Physics - Statistical Mechanics (3)
 
Physics - Strongly Correlated Electrons (3)
 
Physics - Atomic Physics (3)
 
Computer Science - Neural and Evolutionary Computing (3)
 
Mathematics - Optimization and Control (3)
 
Physics - Mesoscopic Systems and Quantum Hall Effect (3)
 
Physics - Geophysics (2)
 
Computer Science - Information Retrieval (2)
 
Physics - Disordered Systems and Neural Networks (2)
 
Computer Science - Computer Vision and Pattern Recognition (2)
 
Computer Science - Networking and Internet Architecture (2)
 
Physics - Computational Physics (2)
 
Physics - Superconductivity (2)
 
Physics - Biological Physics (1)
 
High Energy Physics - Theory (1)
 
Mathematics - Numerical Analysis (1)
 
Mathematical Physics (1)
 
Computer Science - Numerical Analysis (1)
 
Statistics - Computation (1)
 
Mathematics - Algebraic Geometry (1)
 
Mathematics - Mathematical Physics (1)
 
Nuclear Experiment (1)
 
Physics - Optics (1)
 
Mathematics - Information Theory (1)
 
Mathematics - General Topology (1)
 
Mathematics - Combinatorics (1)
 
Mathematics - Algebraic Topology (1)
 
Physics - Instrumentation and Detectors (1)
 
Mathematics - Rings and Algebras (1)
 
Computer Science - Information Theory (1)
 
High Energy Astrophysical Phenomena (1)
 
Nuclear Theory (1)
 
High Energy Physics - Lattice (1)
 
Physics - Fluid Dynamics (1)

Publications Authored By Z. -T. Lu

We demonstrate an electrically induced, non-volatile, metal-insulator phase transition in a MoS$_2$ transistor. A single crystalline, epitaxially grown, PbZr$_{0.2}$Ti$_{0. Read More

By using first-principles calculation, the interaction of O2 and H2O molecules with the pristine and the defective InSe monolayers is studied. It is predicted that the single Se and In vacancies exhibit significantly enhanced chemical activity toward the adsorbates compared with the perfect InSe lattice site, and the Se vacancies have a much higher chemical activity than the In vacancies. H2O molecule should be only physisorbed on the various InSe monolayers at ambient conditions, according to the calculated energies. Read More

The chest X-ray is one of the most commonly accessible radiological examinations for screening and diagnosis of many lung diseases. A tremendous number of X-ray imaging studies accompanied by radiological reports are accumulated and stored in many modern hospitals' Picture Archiving and Communication Systems (PACS). On the other side, it is still an open question how this type of hospital-size knowledge database containing invaluable imaging informatics (i. Read More

Deep Neural Networks (DNNs) have provably enhanced the state-of-the-art Neural Machine Translation (NMT) with their capability in modeling complex functions and capturing complex linguistic structures. However NMT systems with deep architecture in their encoder or decoder RNNs often suffer from severe gradient diffusion due to the non-linear recurrent activations, which often make the optimization much more difficult. To address this problem we propose novel linear associative units (LAU) to reduce the gradient propagation length inside the recurrent unit. Read More

We demonstrate non-volatile, n-type, back-gated, MoS$_{2}$ transistors, placed directly on an epitaxial grown, single crystalline, PbZr$_{0.2}$Ti$_{0.8}$O$_{3}$ (PZT) ferroelectric. Read More

We consider the late-time tailing in a tracer test performed with a push-drift methodology (i.e., quasi-radial injection followed by drift under natural gradient). Read More

We report infrared magneto-spectroscopy studies on thin crystals of an emerging Dirac material ZrTe5 near the intrinsic limit. The observed structure of the Landau level transitions and zero-field infrared absorption indicate a two-dimensional Dirac-like electronic structure, similar to that in graphene but with a small relativistic mass corresponding to a 9.4 meV energy gap. Read More

The electronic structure of recently discovered $\beta$-Fe$_4$Se$_5$ with $\sqrt{5} \times \sqrt{5}$ ordered Fe vacancies is calculated using first-principles density functional theory. We find that the ground state is an antiferromagnetic (AFM) insulator in agreement with the experimental observation. In K$_2$Fe$_4$Se$_5$, it is known that the ground state is $\sqrt{5} \times \sqrt{5}$-blocked-checkerboard AFM ordered. Read More

In this article, we present a qualitative approach to study the dynamics and stability of micro-machined inductive contactless suspensions (MIS). In the framework of this approach, the induced eddy current into a levitated micro-object is considered as a collection of m-eddy current circuits. Assuming small displacements and the quasi- static behaviour of the levitated micro-object, a generalized model of MIS is obtained and represented as a set of six linear differential equations corresponding to six degrees of freedom in a rigid body by using Lagrange-Maxwell formalism. Read More

In this paper we state a problem on rigidity of powers and give a solution of this problem for m=2. Our statement of this problem is elementary enough and does not require any knowledge of algebraic topology. Actually, this problem is related to unitary circle actions, rigid Hirzebruch genera and Kosniowski's conjecture. Read More

We calculate the two leading-twist transverse momentum dependent distribution functions of the pion meson, the unpolarized distribution $f_{1\pi}(x,\bm{k}^2_T)$ and the Boer-Mulders function $h_{1\pi}^\perp (x,\bm{k}^2_T)$, using the pion wave functions derived from a light-cone approach. The evolution effect of the first $\bm k_T$-moment of the pion Boer-Mulders function is studied by employing an approximate evolution kernel. Using the model resulting distributions, we predict the transverse momentum weighted $\cos2\phi$ azimuthal asymmetry in the unpolarized $\pi^- p$ Drell-Yan process which can be measured at COMPASS in the near future. Read More

A large number of application problems involve two levels of optimization, where one optimization task is nested inside the other. These problems are known as bilevel optimization problems and have been studied by both classical optimization community and evolutionary optimization community. Most of the solution procedures proposed until now are either computationally very expensive or applicable to only small classes of bilevel optimization problems adhering to mathematically simplifying assumptions. Read More

Analyses for $^{81}$Kr and noble gases on groundwater from the deepest aquifer system of the Baltic Artesian Basin (BAB) were performed to determine groundwater ages and uncover the flow dynamics of the system on a timescale of several hundred thousand years. We find that the system is controlled by mixing of three distinct water masses: Interglacial or recent meteoric water $(\delta^{18}\text{O} \approx -10.4\unicode{x2030})$ with a poorly evolved chemical and noble gas signature, glacial meltwater $(\delta^{18}\text{O} \leq -18\unicode{x2030})$ with elevated noble gas concentrations, and an old, high-salinity brine component $(\delta^{18}\text{O} \geq -4. Read More

We study large-scale kernel methods for acoustic modeling in speech recognition and compare their performance to deep neural networks (DNNs). We perform experiments on four speech recognition datasets, including the TIMIT and Broadcast News benchmark tasks, and compare these two types of models on frame-level performance metrics (accuracy, cross-entropy), as well as on recognition metrics (word/character error rate). In order to scale kernel methods to these large datasets, we use the random Fourier feature method of Rahimi and Recht (2007). Read More

The Lorentz-invariance-violating Weyl and Dirac fermions have recently attracted intensive interests as new types of particles beyond high-energy physics, and they demonstrate novel physical phenomena such as angle-dependent chiral anomaly and topological Lifshitz transition. Here we predict the existence of Lorentz-invariance-violating Dirac fermions in the YPd$_2$Sn class of Heusler alloys that emerge at the boundary between the electron-like and hole-like pockets in the Brillouin zone, based on the first-principles electronic structure calculations. In combination with the fact that this class of materials was all reported to be superconductors, the YPd$_2$Sn class provides an appropriate platform for studying exotic physical properties distinguished from conventional Dirac fermions, especially for realizing possible topological superconductivity. Read More

Monolayer transition metal dichalcogenide (TMDC) crystals, as direct-gap materials with unusually strong light-matter interaction, have attracted much recent attention. In contrast to the initial understanding, the minima of the conduction band are predicted to be spin split. Because of this splitting and the spin-polarized character of the valence bands, the lowest-lying excitonic states in WX2 (X=S, Se) are expected to be spin-forbidden and optically dark. Read More

Random backpropagation (RBP) is a variant of the backpropagation algorithm for training neural networks, where the transpose of the forward matrices are replaced by fixed random matrices in the calculation of the weight updates. It is remarkable both because of its effectiveness, in spite of using random matrices to communicate error information, and because it completely removes the taxing requirement of maintaining symmetric weights in a physical neural system. To better understand random backpropagation, we first connect it to the notions of local learning and the learning channel. Read More

Building neural networks to query a knowledge base (a table) with natural language is an emerging research topic in NLP. The neural enquirer typically necessitates multiple steps of execution because of the compositionality of queries. In previous studies, researchers have developed either distributed enquirers or symbolic ones for table querying. Read More

We study the production of polarized $\Lambda$ hyperon in semi-inclusive deep inelastic scattering off an unpolarized target. We include the cases in which the $\Lambda$ hyperon is longitudinally polarized or transversely polarized, and in which the lepton beam is unpolarized or longitudinally polarized. Within the framework of the transverse momentum dependent factorization, we take into account the complete decomposition of the parton correlator for fragmentation up to twist-3. Read More

We calculate the gluon Sivers function of the proton in the valence-$x$ region using a light-cone spectator model with the presence of the gluon degree of freedom. We obtain the values of the parameters by fitting the model resulting gluon density distribution to the known parametrization. We find that our results agree with the recent phenomenological extraction of the gluon Sivers function after considering the evolution effect. Read More

Conventional attention-based Neural Machine Translation (NMT) conducts dynamic alignment in generating the target sentence. By repeatedly reading the representation of source sentence, which keeps fixed after generated by the encoder (Bahdanau et al., 2015), the attention mechanism has greatly enhanced state-of-the-art NMT. Read More

Neural Machine Translation (NMT) is a new approach to machine translation that has made great progress in recent years. However, recent studies show that NMT generally produces fluent but inadequate translations (Tu et al. 2016b; Tu et al. Read More

In this work, we study the stochastic alternating direction method of multipliers (ADMM) for the nonconvex optimizations, and propose three classes of nonconvex stochastic ADMM with variance reduction, based on different reduced variance stochastic gradients. Specifically, the first class called the nonconvex stochastic variance reduced gradient ADMM (SVRG-ADMM), uses a multi-stage scheme to progressively reduce the variance of stochastic gradients. The second is the nonconvex stochastic average gradient ADMM (SAG-ADMM), which additionally uses the old gradients estimated in the previous iteration. Read More

In mammography, the efficacy of computer-aided detection methods depends, in part, on the robust localisation of micro-calcifications ($\mu$C). Currently, the most effective methods are based on three steps: 1) detection of individual $\mu$C candidates, 2) clustering of individual $\mu$C candidates, and 3) classification of $\mu$C clusters. Where the second step is motivated both to reduce the number of false positive detections from the first step and on the evidence that malignancy depends on a relatively large number of $\mu$C detections within a certain area. Read More

Studying thermal transport at the nanoscale poses formidable experimental challenges due both to the physics of the measurement process and to the issues of accuracy and reproducibility. The laser-induced transient thermal grating (TTG) technique permits non-contact measurements on nanostructured samples without a need for metal heaters or any other extraneous structures, offering the advantage of inherently high absolute accuracy. We present a review of recent studies of thermal transport in nanoscale silicon membranes using the TTG technique. Read More

In this paper, we study one-Lee weight and two-Lee weight codes over $\mathbb{Z}_{2}\mathbb{Z}_{2}[u]$, where $u^{2}=0$. Some properties of one-Lee weight $\mathbb{Z}_{2}\mathbb{Z}_{2}[u]$-additive codes are given, and a complete classification of one-Lee weight $\mathbb{Z}_2\mathbb{Z}_2[u]$-additive formally self-dual codes is obtained. The structure of two-Lee weight projective $\mathbb{Z}_2\mathbb{Z}_2[u]$ codes are determined. Read More

Under certain conditions, it takes a shorter time to cool a hot system than to cool the same system initiated at a lower temperature. This phenomenon - the "Mpemba Effect" - is well known in water, and has recently been observed in other systems as well. However, there is no single generic mechanism that explains this counter-intuitive behavior. Read More

Al$^+$ ions optical clock is a very promising optical frequency standard candidate due to its extremely small blackbody radiation shift. It has been successfully demonstrated with indirect cooled, quantum-logic-based spectroscopy technique. Its accuracy is limited by second-order Doppler shift, and its stability is limited by the number of ions that can be probed in quantum logic processing. Read More

We study the influence of external electric, $E$, and magnetic, $B$, fields parallel to each other, and of a chiral chemical potential, $\mu_5$, on the chiral phase transition of Quantum Chromodynamics. Our theoretical framework is a Nambu-Jona-Lasinio model with a contact interaction. Within this model we compute the critical temperature of chiral symmetry restoration, $T_c$, as a function of the chiral chemical potential and field strengths. Read More

We report the "Recurrent Deterioration" (RD) phenomenon observed in online recommender systems. The RD phenomenon is reflected by the trend of performance degradation when the recommendation model is always trained based on users' feedbacks of the previous recommendations. There are several reasons for the recommender systems to encounter the RD phenomenon, including the lack of negative training data and the evolution of users' interests, etc. Read More

In neural machine translation (NMT), generation of a target word depends on both source and target contexts. We find that source contexts have a direct impact on the adequacy of a translation while target contexts affect the fluency. Intuitively, generation of a content word should rely more on the source context and generation of a functional word should rely more on the target context. Read More

The out-of-time-ordered (OTO) correlation is a key quantity for quantifying quantum chaoticity and has been recently used in the investigation of quantum holography. Here we use it to study and characterize many-body localization (MBL). We find that a long-time logarithmic variation of the OTO correlation occurs in the MBL phase but is absent in the Anderson localized and ergodic phases. Read More

The emission from black hole binaries (BHBs) and active galactic nuclei (AGNs) displays significant aperiodic variabilities. The most promising explanation for these variabilities is the propagating fluctuations in the accretion flow. It is natural to expect that the mechanism driving variabilities in BHBs and AGNs may operate in a black hole hyper-accretion disk, which is believed to power gamma-ray bursts (GRBs). Read More

The main approach of traditional information retrieval (IR) is to examine how many words from a query appear in a document. A drawback of this approach, however, is that it may fail to detect relevant documents where no or only few words from a query are found. The semantic analysis methods such as LSA (latent semantic analysis) and LDA (latent Dirichlet allocation) have been proposed to address the issue, but their performance is not superior compared to common IR approaches. Read More

We present a calculation of the twist-3 T-odd chiral-even fragmentation functions $G^{\perp}$ and $\tilde{G}^{\perp}$ using a spectator model. We consider the effect gluon exchange to calculate all necessary one-loop diagrams for the quark-quark and quark-gluon-quark correlation functions. We find that the gluon loops corrections generate non-zero contribution to these two fragmentation function. Read More

We propose a numerical method for explicitly constructing a complete set of local integrals of motion (LIOM) and definitely show the existence of LIOM for strongly many-body localized systems. The method combines exact diagonalization and nonlinear minimization, and gradually deforms the LIOM for the noninteracting case to those for the interacting case. By using this method we find that for strongly disordered and weakly interacting systems, there are two characteristic lengths in the LIOM. Read More

In this paper we consider the composite self-concordant (CSC) minimization problem, which minimizes the sum of a self-concordant function $f$ and a (possibly nonsmooth) proper closed convex function $g$. The CSC minimization is the cornerstone of the path-following interior point methods for solving a broad class of convex optimization problems. It has also found numerous applications in machine learning. Read More

Symmetry protected topological (SPT) phases in free fermion and interacting bosonic systems have been classified, but the physical phenomena of interacting fermionic SPT phases have not been fully explored. Here, employing large-scale quantum Monte Carlo simulation, we investigate the edge physics of a bilayer Kane-Mele-Hubbard model with zigzag ribbon geometry. Our unbiased numerical results show that the fermion edge modes are gapped out by interaction, while the bosonic edge modes remain gapless at the $(1+1)d$ boundary, before the bulk quantum phase transition to a topologically trivial phase. Read More

Background: Octupole-deformed nuclei, such as that of $^{225}$Ra, are expected to amplify observable atomic electric dipole moments (EDMs) that arise from time-reversal and parity-violating interactions in the nuclear medium. In 2015, we reported the first "proof-of-principle" measurement of the $^{225}$Ra atomic EDM. Purpose: This work reports on the first of several experimental upgrades to improve the statistical sensitivity of our $^{225}$Ra EDM measurements by orders of magnitude and evaluates systematic effects that contribute to current and future levels of experimental sensitivity. Read More

In this paper, we investigate how to network smartphones for providing communications in disaster recovery. By bridging the gaps among different kinds of wireless networks, we have designed and implemented a system called TeamPhone, which provides smartphones the capabilities of communications in disaster recovery. Specifically, TeamPhone consists of two components: a messaging system and a self-rescue system. Read More

Opportunistic mobile networks consisting of intermittently connected mobile devices have been exploited for various applications, such as computational offloading and mitigating cellular traffic load. In contrast to existing work, in this paper, we focus on cooperatively offloading data among mobile devices to maximally improve the probability of data delivery from a mobile device to intermittently connected infrastructure within a given time constraint, which is referred to as the \textit{cooperative offloading} problem. Unfortunately, the estimation of data delivery probability over an opportunistic path is difficult and cooperative offloading is NP-hard. Read More

We study the longitudinal-transverse double-spin asymmetry with a $\cos\phi_S$ modulation in semi-inclusive deep inelastic scattering for charged and neutral pions production. We consider the particular case in which the transverse momentum of the final state hadron is integrated out. The corresponding asymmetry may be contributed by two parts: one is the convolution of the twist-3 distribution function $g_{T}(x)$ and the unpolarized fragmentation function $D_1(z)$, the other is related to the coupling of the transversity distribution function $h_1(x)$ and the collinear twist-3 fragmentation function $\tilde{E}(z)$. Read More

In this paper, we propose phraseNet, a neural machine translator with a phrase memory which stores phrase pairs in symbolic form, mined from corpus or specified by human experts. For any given source sentence, phraseNet scans the phrase memory to determine the candidate phrase pairs and integrates tagging information in the representation of source sentence accordingly. The decoder utilizes a mixture of word-generating component and phrase-generating component, with a specifically designed strategy to generate a sequence of multiple words all at once. Read More

We propose to enhance the RNN decoder in a neural machine translator (NMT) with external memory, as a natural but powerful extension to the state in the decoding RNN. This memory-enhanced RNN decoder is called \textsc{MemDec}. At each time during decoding, \textsc{MemDec} will read from this memory and write to this memory once, both with content-based addressing. Read More

Efficient fiber-to-waveguide light coupling has been a key issue in integrated photonics for many years. The main challenge lies in the huge mode mismatch between an optical fiber and a single mode waveguide. Herein, we present a novel fiber-to-waveguide coupler, named "L-coupler", through which the light fed from the top of a chip can bend 90{\deg} with low reflection and is then efficiently coupled into an on-chip Si waveguide within a short propagation distance (<20{\mu}m). Read More

Let the vector bundle $\mathcal{E}$ be a deformation of the tangent bundle over the Grassmannian $G(k,n)$. We compute the ring structure of sheaf cohomology valued in exterior powers of $\mathcal{E}$, also known as the polymology. This is the first part of a project studying the quantum sheaf cohomology of Grassmannians with deformations of the tangent bundle, a generalization of ordinary quantum cohomology rings of Grassmannians. Read More

Recurrent neural networks (RNNs), including long short-term memory (LSTM) RNNs, have produced state-of-the-art results on a variety of speech recognition tasks. However, these models are often too large in size for deployment on mobile devices with memory and latency constraints. In this work, we study mechanisms for learning compact RNNs and LSTMs via low-rank factorizations and parameter sharing schemes. Read More

By using the first-principles electronic structure calculations, we have systematically studied the magnetism in three recently synthesized iron-based antiperovskite chalco-halides: Ba$_3$(FeS$_4$)Cl, Ba$_3$(FeS$_4$)Br, and Ba$_3$(FeSe$_4$)Br. These compounds consist of edge-sharing Ba$Q_6$ ($Q$=Cl or Br) octahedra intercalated with isolated Fe$X_4$ ($X$=S or Se) tetrahedra. We find that even though the shortest distances between the nearest-neighboring Fe atoms in these three compounds already exceed 6 \AA, much larger than the bond length of a chemical bonding, they all remarkably show antiferromagnetic (AFM) coupling along $b$ axis with very weak spin-spin correlation along $a$ axis. Read More