arXiv subjects

Jing Zeng

Publications and source records attributed to Jing Zeng.

30 records · Page 2Linked to original sources

arXiv2021

Boost-S: Gradient Boosted Trees for Spatial Data and Its Application to FDG-PET Imaging Data

Boosting Trees are one of the most successful statistical learning approaches that involve sequentially growing an ensemble of simple regression trees (i.e., "weak learners"). However, gradient boosted trees are not yet available for spatially correlated data. This paper proposes a new gradient Boosted Trees algorithm for Spatial Data (Boost-S) with covariate information. Boost-S integrates the spatial correlation structure into the classical framework of gradient boosted trees. Each tree is grown by solving a regularized optimization problem, where the objective function involves two penalty terms on tree complexity and takes into account the underlying spatial correlation. A computationally-efficient algorithm is proposed to obtain the ensemble trees. The proposed Boost-S is applied to the spatially-correlated FDG-PET (fluorodeoxyglucose-positron emission tomography) imaging data collected during cancer chemoradiotherapy. Our numerical investigations successfully demonstrate the advantages of the proposed Boost-S over existing approaches for this particular application.

stat.AP↗

arXiv2018

Spontaneous emission of a moving atom in a waveguide of rectangular cross section

We study the spontaneous emission (SE) of an excited two-level nonrelativistic system (TLS) interacting with the vacuum in a waveguide of rectangular cross section. All TLS's transitions and the center-of-mass motion of the TLS are taken into account. The SE rate and the carried frequency of the emitted photon for the TLS initial being at rest is obtained, it is found in the first order of the center of mass (c.m.) that the frequency of the emitted photon could be smaller or larger than the transition frequency of the TLS but the SE rate is smaller than the SE rate $Γ_{f}$ of the TLS fixed in the same waveguide. The SE rate and the carried frequency of the emitted photon for the TLS initial being moving is also obtained in the first order of the c.m.. The SE rate is larger than $Γ_{f}$ but it is independent of the initial momentum. The carried frequency of the emitted photon is creased when it travels along the direction of the initial momentum and is decreased when it travels in the opposite direction of the initial momentum.

quant-ph↗

arXiv2016

Band gap modulation in γ-graphyne by p-n codoping

The modulation of band gap in the two-dimensional carbon materials is of impor- tance for their applications as electronic devices. By first-principles calculations, we propose a model to control the band gap size of γ-graphyne. The model is named as p-n codoping, i. e., using B and N atoms to codope into γ-graphyne. After codoping, B atom plays a role of p doping and N atom acts as n doping. The Fermi energy level returns around the forbidden zone and the band gap of γ-graphyne vary bigger or smaller. Moreover, the gaps exhibit an oscillated behaviour in different codoping configurations. The proposed model serves as new insights for better modulation of the electronic properties of 2D carbon materials.

cond-mat.mes-hall↗

arXiv2016

The mechanism and process of spontaneous boron doping in graphene in the theoretical perspective

A theoretical model is presented to reveal the mechanism of B doping into graphene in the microwave plasma experiment choosing trimethylboron as the doping source (ACS NANO 6 (2012) 1970). The results show that the reason for B doping comes from the combinational interaction of B and other groups (C, H, CH, CH2 or CH3) decomposing from trimethylboron and the doping undergoes two crucial steps. The minimal energy path for the first step are determined. The obtained energy barrier of considered cases fall into the range of 0.02-0.43 eV, supporting the fact that the substituting B for C can easily realized even at room temperature. As the second step, after removing irrelevant groups in vertical direction through H saturation, the perfect B doping is realized at last. This work successfully explain the above experimental phenomenon and propose a novel and feasible method aiming at B doping of graphene.

cond-mat.mes-hall↗

arXiv2015

Liouville type theorems for stable solutions of the weighted elliptic system

We examine the weighted elliptic system \begin{equation*} \begin{cases} -Δu=(1+|x|^2)^{\fracα{2}} v,\\ -Δv=(1+|x|^2)^{\fracα{2}} u^p, \end{cases} \quad \mbox{in}\;\ \mathbb{R}^N, \end{equation*}where $N \ge 5$, $p>1$ and $α>0$. We prove Liouville type results for the classical positive (nonnegative) stable solutions in dimension $N<\ell+\dfrac{α(\ell-2)}{2}$ ($N <\ell+\dfrac{α(\ell-2)(p+3)}{4(p+1)}$) and $\ell \ge 5$, $p \in (1,p_*(\ell))$. In particular, for any $p>1$ and $α> 0$, we obtain the nonexistence of classical positive (nonnegative) stable solutions for any $N \le 12+5 α$ ($N\le 12+\dfrac{5α(p+3)}{2(p+1)}$).

math.AP↗

arXiv2015

A Computer-Assisted Stability Proof for a Stationary Solution of Reaction-Diffusion Equations

The main subject of this paper is a computer assisted stability proof for a stationary solution of reaction diffusion equations in one dimensional space. We use Nakao's numerical verification method to enclose a stationary solution of reaction-diffusion equations. Considering the linearized stability of the solution, a method of excluding eigenvalues in a half plane is adopted. We first focus on the eigenvalues for an operator linearized at an approximate solution. An excluding theorem is presented such that we know under some condition, and there is no eigenvalue in some disks. Some computable criteria are constructed to apply the theorem in a computer. And also the invertibility of some operator is proved theoretically in the paper. However, we need the information of the eigenvalues for the operator linearized at the exact solution. This can be obtained by combining with the verification results of the solution. Then we judge the stability of the solution from the domain where the eigenvalues are located. At last there are some verification results.

math.NA↗

arXiv2014

Infinite solutions having a prescribed number of nodes for a p-Laplacian problem

In this paper, we are concern with the multiplicity of solutions for a p-Laplacian problem. A weaker super-quadratic assumptions is required on the nonlinearity. Under the weaker condition we give a new proof for the infinite solutions having a prescribed number of nodes to the problem. It turns out that the weaker condition on nonlinearity suffices to guarantee the infinitely many solutions. At the same time, a global characterization of the critical values of the nodal radial solutions are given.

math.AP↗

arXiv2014

Vector rectangular-shape laser based on reduced graphene oxide interacting with long fiber taper

A vector dual-wavelength rectangular-shape laser (RSL) based on a long fiber taper deposited with reduced graphene oxide is proposed, where the nonlinearity is enhanced due to large evanescent-field-interacting length and strong field confinement of a 8 mm fiber taper with a waist diameter of 4 micronmeters. Graphene flakes are deposited uniformly on the taper waist with light pressure effect, so this structure guarantees both excellent saturable absorption and high nonlinearity. The RSL with a repetition rate of 7.9 MHz exhibits fast polarization switching in two orthogonal polarization directions, and the temporal and spectral characteristics are investigated. The results suggest that the long taper-based graphene structure is an efficient choice for nonlinear devices.

physics.optics↗

arXiv2014

Wavelength Spacing Tunable, Multiwavelength Q-Switched Mode-Locked Laser Based on Graphene-Oxide-Deposited Tapered Fiber

A wavelength spacing tunable, multiwavelength Q-switched mode-locked (QML) fiber laser in an erbium-doped fiber cavity based on graphene oxide deposited on tapered fiber is proposed by choosing the diameter and length of the taper, graphene oxide thickness and cavity dispersion, in which the wavelength spacing could be tuned by pump power. The evolutions of temporal and spectral with different pump strengths are investigated. Results show that the tunability of the multiwavelength laser can be interpreted by the bound states of QML laser resulting from a mutual interaction of dispersion, nonlinear effect, insertion loss, and pump power. To the best of our knowledge, it is the first experimental observation of bound states of QML, which provides a new mechanism to fabricate tunable multiwavelength laser.

physics.optics↗

arXiv2012

An optimal bound on the number of interior spike solutions for Lin-Ni-Takagi problem

We consider the following singularly perturbed Neumann problem {eqnarray*} \ve^2 Δu -u +u^p = 0 \quad {in} \quad Ω, \quad u>0 \quad {in} \quad Ω, \quad {\partial u \over \partial ν}=0 \quad {on} \quad \partial Ω, {eqnarray*} where $p$ is subcritical and $Ω$ is a smooth and bounded domain in $\R^n$ with its unit outward normal $ν$. Lin-Ni-Wei \cite{LNW} proved that there exists $\ve_0$ such that for $0<\ve<\ve_0$ and for each integer $k$ bounded by {equation} 1\leq k\leq \frac{δ(Ω,n,p)}{(\ve |\log \ve |)^n} {equation} where $δ(Ω,n,p)$ is a constant depending only on $Ω$, $p$ and $n$, there exists a solution with $k$ interior spikes. We show that the bound on $k$ can be improved to {equation} 1\leq k\leq \frac{δ(Ω,n,p)}{\ve^n}, {equation} which is optimal.

math.AP↗

arXiv2012

Leading-order spin-orbit and spin(1)-spin(2) radiation-reaction Hamiltonians

In the present paper, the leading-order post-Newtonian spin-orbit and spin(1)-spin(2) radiation-reaction Hamiltonians are calculated. We utilize the canonical formalism of Arnowitt, Deser, and Misner (ADM), which has shown to be valuable for this kind of calculation. The results are valid for arbitrary many objects. The energy loss is then computed and compared to well-known results for the energy flux as a check.

gr-qc↗

arXiv2007

Application of energy and angular momentum balance to gravitational radiation reaction for binary systems with spin-orbit coupling

We study gravitational radiation reaction in the equations of motion for binary systems with spin-orbit coupling, at order (v/c)^7 beyond Newtonian gravity, or O(v/c)^2 beyond the leading radiation reaction effects for non-spinning bodies. We use expressions for the energy and angular momentum flux at infinity that include spin-orbit corrections, together with an assumption of energy and angular momentum balance, to derive equations of motion that are valid for general orbits and for a class of coordinate gauges. We show that the equations of motion are compatible with those derived earlier by a direct calculation.

gr-qc↗