arXiv subjects

Hong Zhang

Publications and source records attributed to Hong Zhang.

At least 307 records · Page 17Linked to original sources

arXiv2017

Pareto-efficient biological pest control enable high efficacy at small costs

Biological pest control is increasingly used in agriculture as a an alternative to traditional chemical pest control. In many cases, this involves a one-off or periodic release of naturally occurring and/or genetically modified enemies such as predators, parasitoids, or pathogens. As the interaction between these enemies and the pest is complex and the production of natural enemies potentially expensive, it is not surprising that both the efficacy and economic viability of biological pest control are debated. Here, we investigate the performance of very simple control strategies. In particular, we show how Pareto-efficient one-off or periodic release strategies, that optimally trade off between efficacy and economic viability, can be devised and used to enable high efficacy at small economic costs. We demonstrate our method on a pest-pathogen-crop model with a tunable immigration rate of pests. By analyzing this model, we demonstrate that simple Pareto-efficient one-off and periodic release strategies are efficacious and simultaneously have profits that are close to the theoretical maximum obtained by strategies optimizing only the profit. When the immigration rate of pests is low to intermediate, one-off control strategies are sufficient, and when the immigration of pests is high, periodic release strategies are preferable. The methods presented here can be extended to more complex scenarios and be used to identify promising biological pest control strategies in many circumstances.

q-bio.PE↗

arXiv2017

Mass Dependence of Higgs Production at Large Transverse Momentum through a Bottom Quark Loop

In the production of the Higgs through a bottom-quark loop, the transverse momentum distribution of the Higgs at large $P_T$ is complicated by its dependence on two other important scales: the bottom quark mass $m_b$ and the Higgs mass $m_H$. A strategy for simplifying the calculation of the cross section at large $P_T$ is to calculate only the leading terms in its expansion in $m_b^2/P_T^2$. In this paper, we consider the bottom-quark-loop contribution to the parton process $q\bar{q}\to H+g$ at leading order in $α_s$. We show that the leading power of $1/P_T^2$ can be expressed in the form of a factorization formula that separates the large scale $P_T$ from the scale of the masses. All the dependence on $m_b$ and $m_H$ can be factorized into a distribution amplitude for $b \bar b$ in the Higgs, a distribution amplitude for $b \bar b$ in a real gluon, and an endpoint contribution. The factorization formula can be used to organize the calculation of the leading terms in the expansion in $m_b^2/P_T^2$ so that every calculation involves at most two scales.

hep-ph↗

arXiv2017

$J/ψ$ production and polarization within a jet

We study the production and polarization of $J/ψ$ mesons within a jet in proton-proton collisions at the LHC. We define the $J/ψ$-jet fragmentation function as a ratio of differential jet cross sections with and without the reconstructed $J/ψ$ in the jet. We demonstrate that this is a very useful observable to help explore the $J/ψ$ production mechanism, and to differentiate between different NRQCD global fits based on inclusive $J/ψ$ cross sections. Furthermore, we propose to measure the polarization of $J/ψ$ mesons inside the jet, which can provide even more stringent constraints for the heavy quarkonium production mechanism.

hep-ph↗

arXiv2017

Convolutional Neural Network-Based Block Up-sampling for Intra Frame Coding

Inspired by the recent advances of image super-resolution using convolutional neural network (CNN), we propose a CNN-based block up-sampling scheme for intra frame coding. A block can be down-sampled before being compressed by normal intra coding, and then up-sampled to its original resolution. Different from previous studies on down/up-sampling-based coding, the up-sampling methods in our scheme have been designed by training CNN instead of hand-crafted. We explore a new CNN structure for up-sampling, which features deconvolution of feature maps, multi-scale fusion, and residue learning, making the network both compact and efficient. We also design different networks for the up-sampling of luma and chroma components, respectively, where the chroma up-sampling CNN utilizes the luma information to boost its performance. In addition, we design a two-stage up-sampling process, the first stage being within the block-by-block coding loop, and the second stage being performed on the entire frame, so as to refine block boundaries. We also empirically study how to set the coding parameters of down-sampled blocks for pursuing the frame-level rate-distortion optimization. Our proposed scheme is implemented into the High Efficiency Video Coding (HEVC) reference software, and a comprehensive set of experiments have been performed to evaluate our methods. Experimental results show that our scheme achieves significant bits saving compared with HEVC anchor especially at low bit rates, leading to on average 5.5% BD-rate reduction on common test sequences and on average 9.0% BD-rate reduction on ultra high definition (UHD) test sequences.

cs.MM↗

arXiv2017

Scalar curvature flow on S^n to a prescribed sign-changing function

In this paper, we consider the problem of prescribing scalar curvature on n-sphere. Assume that the candidate curvature function $f$, which is allowed to change sign, satisfies some kind of Morse index or symmetry condition. By studying the well-known scalar curvature flow, we are able to prove that the flow converges to a metric with the prescribed function $f$ as its scalar curvature.

math.AP↗

arXiv2017

Mass Dependence of Higgs Production at Large Transverse Momentum

The transverse momentum distribution of the Higgs at large $P_T$ is complicated by its dependence on three important energy scales: $P_T$, the top quark mass $m_t$, and the Higgs mass $m_H$. A strategy for simplifying the calculation of the cross section at large $P_T$ is to calculate only the leading terms in its expansion in $m_t^2/P_T^2$ and/or $m_H^2/P_T^2$. The expansion of the cross section in inverse powers of $P_T$ is complicated by logarithms of $P_T$ and by mass singularities. In this paper, we consider the top-quark loop contribution to the subprocess $q\bar{q}\to H+g$ at leading order in $α_s$. We show that the leading power of $1/P_T^2$ can be expressed in the form of a factorization formula that separates the large scale $P_T$ from the scale of the masses. All the dependence on $m_t$ and $m_H$ can be factorized into a distribution amplitude for $t \bar t$ in the Higgs, a distribution amplitude for $t \bar t$ in a real gluon, and an endpoint contribution. The factorization formula can be used to simplify calculations of the $P_T$ distribution at large $P_T$ to next-to-leading order in $α_s$.

hep-ph↗

arXiv2017

On Schauder estimates for a class of nonlocal fully nonlinear parabolic equations

We obtain Schauder estimates for a class of concave fully nonlinear nonlocal parabolic equations of order $σ\in (0,2)$ with rough and non-symmetric kernels. As a application, we prove that the solution to a translation invariant equation with merely bounded data is $C^σ$ in $x$ variable and $Λ^1$ in $t$ variable, where $Λ^1$ is the Zygmund space.

math.AP↗

arXiv2017

On The Stability of Video Detection and Tracking

In this paper, we study an important yet less explored aspect in video detection and tracking -- stability. Surprisingly, there is no prior work that tried to study it. As a result, we start our work by proposing a novel evaluation metric for video detection which considers both stability and accuracy. For accuracy, we extend the existing accuracy metric mean Average Precision (mAP). For stability, we decompose it into three terms: fragment error, center position error, scale and ratio error. Each error represents one aspect of stability. Furthermore, we demonstrate that the stability metric has low correlation with accuracy metric. Thus, it indeed captures a different perspective of quality. Lastly, based on this metric, we evaluate several existing methods for video detection and show how they affect accuracy and stability. We believe our work can provide guidance and solid baselines for future researches in the related areas.

cs.CV↗

arXiv2017

Nekrasov and Argyres-Douglas theories in spherical Hecke algebra representation

AGT conjecture connects Nekrasov instanton partition function of 4D quiver gauge theory with 2D Liouville conformal blocks. We re-investigate this connection using the central extension of spherical Hecke algebra in q-coordinate representation, q being the instanton expansion parameter. Based on AFLT basis together with interwiners we construct gauge conformal state and demonstrate its equivalence to the Liouville conformal state, with careful attention to the proper scaling behavior of the state. Using the colliding limit of regular states, we obtain the formal expression of irregular conformal states corresponding to Argyres-Douglas theory, which involves summation of functions over Young diagrams.

hep-th↗

arXiv2017

Face Recognition using Multi-Modal Low-Rank Dictionary Learning

Face recognition has been widely studied due to its importance in different applications; however, most of the proposed methods fail when face images are occluded or captured under illumination and pose variations. Recently several low-rank dictionary learning methods have been proposed and achieved promising results for noisy observations. While these methods are mostly developed for single-modality scenarios, recent studies demonstrated the advantages of feature fusion from multiple inputs. We propose a multi-modal structured low-rank dictionary learning method for robust face recognition, using raw pixels of face images and their illumination invariant representation. The proposed method learns robust and discriminative representations from contaminated face images, even if there are few training samples with large intra-class variations. Extensive experiments on different datasets validate the superior performance and robustness of our method to severe illumination variations and occlusion.

cs.CV↗

arXiv2017

Rotation-invariant observables in polarization measurements

Polarization measurements provide a detailed method to test the Standard Model and to search for new physics. Most previous studies depend on pre-selected coordinates, which blurs the significance of the results. The construction of two rotation-invariant observables in vector boson decay into a fermion pair has been proved to be a big success. In this work, we show that there are more rotation-invariant observables and provide a general recipe to find all of them in an arbitrary decay process. Taking spin-1/2 and spin-1 particle decay processes as examples, we calculate the explicit expressions of all rotation-invariant observables, which can serve as a robust test of the detector acceptance and help the analysis of experimental data.

hep-ph↗

arXiv2017

Distributions and Statistical Power of Optimal Signal-Detection Methods In Finite Cases

In big data analysis for detecting rare and weak signals among $n$ features, some grouping-test methods such as Higher Criticism test (HC), Berk-Jones test (B-J), and $ϕ$-divergence test share the similar asymptotical optimality when $n \rightarrow \infty$. However, in practical data analysis $n$ is frequently small and moderately large at most. In order to properly apply these optimal tests and wisely choose them for practical studies, it is important to know how to get the p-values and statistical power of them. To address this problem in an even broader context, this paper provides analytical solutions for a general family of goodness-of-fit (GOF) tests, which covers these optimal tests. For any given i.i.d. and continuous distributions of the input test statistics of the $n$ features, both p-value and statistical power of such a GOF test can be calculated. By calculation we compared the finite-sample performances of asymptotically optimal tests under the normal mixture alternative. Results show that HC is the best choice when signals are rare, while B-J is more robust over various signal patterns. In the application to a real genome-wide association study, results illustrate that the p-value calculation works well, and the optimal tests have potentials for detecting novel disease genes with weak genetic effects. The calculations have been implemented in an R package SetTest and published on the CRAN.

math.ST↗

arXiv2016

Dini estimates for nonlocal fully nonlinear elliptic equations

We obtain Dini type estimates for a class of concave fully nonlinear nonlocal elliptic equations of order $σ\in (0,2)$ with rough and non-symmetric kernels. The proof is based on a novel application of Campanato's approach and a refined $C^{σ+α}$ estimate in [8].

math.AP↗

arXiv2016

Joint Projection and Dictionary Learning using Low-rank Regularization and Graph Constraints

In this paper, we aim at learning simultaneously a discriminative dictionary and a robust projection matrix from noisy data. The joint learning, makes the learned projection and dictionary a better fit for each other, so a more accurate classification can be obtained. However, current prevailing joint dimensionality reduction and dictionary learning methods, would fail when the training samples are noisy or heavily corrupted. To address this issue, we propose a joint projection and dictionary learning using low-rank regularization and graph constraints (JPDL-LR). Specifically, the discrimination of the dictionary is achieved by imposing Fisher criterion on the coding coefficients. In addition, our method explicitly encodes the local structure of data by incorporating a graph regularization term, that further improves the discriminative ability of the projection matrix. Inspired by recent advances of low-rank representation for removing outliers and noise, we enforce a low-rank constraint on sub-dictionaries of all classes to make them more compact and robust to noise. Experimental results on several benchmark datasets verify the effectiveness and robustness of our method for both dimensionality reduction and image classification, especially when the data contains considerable noise or variations.

cs.CV↗

arXiv2016

Object Classification with Joint Projection and Low-rank Dictionary Learning

For an object classification system, the most critical obstacles towards real-world applications are often caused by large intra-class variability, arising from different lightings, occlusion and corruption, in limited sample sets. Most methods in the literature would fail when the training samples are heavily occluded, corrupted or have significant illumination or viewpoint variations. Besides, most of the existing methods and especially deep learning-based methods, need large training sets to achieve a satisfactory recognition performance. Although using the pre-trained network on a generic large-scale dataset and fine-tune it to the small-sized target dataset is a widely used technique, this would not help when the content of base and target datasets are very different. To address these issues, we propose a joint projection and low-rank dictionary learning method using dual graph constraints (JP-LRDL). The proposed joint learning method would enable us to learn the features on top of which dictionaries can be better learned, from the data with large intra-class variability. Specifically, a structured class-specific dictionary is learned and the discrimination is further improved by imposing a graph constraint on the coding coefficients, that maximizes the intra-class compactness and inter-class separability. We also enforce low-rank and structural incoherence constraints on sub-dictionaries to make them more compact and robust to variations and outliers and reduce the redundancy among them, respectively. To preserve the intrinsic structure of data and penalize unfavourable relationship among training samples simultaneously, we introduce a projection graph into the framework, which significantly enhances the discriminative ability of the projection matrix and makes the method robust to small-sized and high-dimensional datasets.

cs.CV↗

arXiv2016

Nonrelativistic Effective Field Theory for Axions

Axions can be described by a relativistic field theory with a real scalar field $ϕ$ whose self-interaction potential is a periodic function of $ϕ$. Low-energy axions, such as those produced in the early universe by the vacuum misalignment mechanism, can be described more simply by a nonrelativistic effective field theory with a complex scalar field $ψ$ whose effective potential is a function of $ψ^*ψ$. We determine the coefficients in the expansion of the effective potential to fifth order in $ψ^*ψ$ by matching low-energy axion scattering amplitudes. In order to describe a Bose-Einstein condensate of axions that is too dense to expand the effective potential in powers of $ψ^*ψ$, we develop a sequence of systematically improvable approximations to the effective potential that include terms of all orders in $ψ^*ψ$.

hep-ph↗

arXiv2016

A moving mesh finite difference method for non-monotone solutions of non-equilibrium equations in porous media

An adaptive moving mesh finite difference method is presented to solve two types of equations with dynamic capillary pressure term in porous media. One is the non-equilibrium Richards Equation and the other is the modified Buckley-Leverett equation. The governing equations are discretized with an adaptive moving mesh finite difference method in the space direction and an implicit-explicit method in the time direction. In order to obtain high quality meshes, an adaptive time-dependent monitor function with directional control is applied to redistribute the mesh grid in every time step, and a diffusive mechanism is used to smooth the monitor function. The behaviors of the central difference flux, the standard local Lax-Friedrich flux and the local Lax-Friedrich flux with reconstruction are investigated by solving a 1D modified Buckley-Leverett equation. With the moving mesh technique, good mesh quality and high numerical accuracy are obtained. A collection of one-dimensional and two-dimensional numerical experiments is presented to demonstrate the accuracy and effectiveness of the proposed method.

physics.comp-ph↗

arXiv2016

A numerical study of two-phase flow models with dynamic capillary pressure and hysteresis

Saturation overshoot and pressure overshoot are studied by incorporating dynamic capillary pressure, capillary pressure hysteresis and hysteretic dynamic coefficient with a traditional fractional flow equation in one dimension. Using the method of lines, the discretizations are constructed by applying Castillo-Grone's mimetic operators in the space direction and explicit trapezoidal integrator in the time direction. Convergence tests and conservation property of the schemes are presented. Computed profiles capture both the saturation overshoot and pressure overshoot phenomena. Comparisons between numerical results and experiments illustrate the effectiveness and different features of the models.

physics.comp-ph↗