arXiv subjects

Fang Su

Publications and source records attributed to Fang Su.

At least 19 recordsLinked to original sources

arXiv2026

Finite-dimensional approximations of random attractor for stochastic discrete complex Ginzburg-Landau equations

In this paper, we apply an implicit Euler scheme to discretize the complex Ginzburg-Landau equation and prove the existence of a numerical attractor for the discrete Ginzburg-Landau system. We establish the upper semicontinuity of the numerical attractor with respect to the global attractor as the time step tends to zero. Furthermore, we provide finite-dimensional approximations for three types of attractors (global, numerical, and random), and demonstrate the existence of truncated attractors along with their convergence as the dimension of the state space tends to infinity. Finally, we prove the existence of a random attractor and establish the upper semi-continuity both of the global random attractor and the truncated random attractor.

math.NA↗

arXiv2026

Approximation for stochastic time-space fractional cable equations driven by rough noise

The time-space fractional cable equation arises from extending the generalized fractional Ohm's law to model anomalous diffusion processes. In this paper, we develop and analyze a numerical approximation for stochastic nonlinear time-space fractional cable equation driven by rough noise. The model involves both two nonlocal terms in time and one in space. By an operator theoretic approach, we establish the existence, uniqueness, and regularities of solutions. We also obtain a convergence result for the regularized equation via Wong-Zakai approximation to regularize the rough noise. The numerical scheme approximates the model in space by the standard spectral Galerkin method and in time by the backward Euler convolution quadrature method. After that, error estimates are established.

math.NA↗

arXiv2025

Approximation of invariant measures for random lattice reversible Selkov systems

This paper focuses on the numerical approximation of random lattice reversible Selkov systems. It establishes the existence of numerical invariant measures for random models with nonlinear noise, using the backward Euler-Maruyama (BEM) scheme for time discretization. The study examines both infinite dimensional discrete random models and their corresponding finite dimensional truncations. A classical path convergence technique is employed to demonstrate the convergence of the invariant measures of the BEM scheme to those of the random lattice reversible Selkov systems. As the discrete time step size approaches zero, the invariant measure of the random lattice reversible Selkov systems can be approximated by the numerical invariant measure of the finite dimensional truncated systems.

math.NA↗

arXiv2025

A TRRIP Down Memory Lane: Temperature-Based Re-Reference Interval Prediction For Instruction Caching

Modern mobile CPU software pose challenges for conventional instruction cache replacement policies due to their complex runtime behavior causing high reuse distance between executions of the same instruction. Mobile code commonly suffers from large amounts of stalls in the CPU frontend and thus starvation of the rest of the CPU resources. Complexity of these applications and their code footprint are projected to grow at a rate faster than available on-chip memory due to power and area constraints, making conventional hardware-centric methods for managing instruction caches to be inadequate. We present a novel software-hardware co-design approach called TRRIP (Temperature-based Re-Reference Interval Prediction) that enables the compiler to analyze, classify, and transform code based on "temperature" (hot/cold), and to provide the hardware with a summary of code temperature information through a well-defined OS interface based on using code page attributes. TRRIP's lightweight hardware extension employs code temperature attributes to optimize the instruction cache replacement policy resulting in the eviction rate reduction of hot code. TRRIP is designed to be practical and adoptable in real mobile systems that have strict feature requirements on both the software and hardware components. TRRIP can reduce the L2 MPKI for instructions by 26.5% resulting in geomean speedup of 3.9%, on top of RRIP cache replacement running mobile code already optimized using PGO.

cs.AR↗

arXiv2025

DEER: Deep Runahead for Instruction Prefetching on Modern Mobile Workloads

Mobile workloads incur heavy frontend stalls due to increasingly large code footprints as well as long repeat cycles. Existing instruction-prefetching techniques suffer from low coverage, poor timeliness, or high cost. We provide a SW/HW co-designed I-prefetcher; DEER uses profile analysis to extract metadata information that allow the hardware to prefetch the most likely future instruction cachelines, hundreds of instructions earlier. This profile analysis skips over loops and recursions to go deeper into the future, and uses a return-address stack on the hardware side to allow prefetch on the return-path from large call-stacks. The produced metadata table is put in DRAM, pointed to by an in-hardware register; the high depth of the lookahead allows to preload the metadata in time and thus nearly no on-chip metadata storage is needed. Gem5 evaluation on real-world modern mobile workloads shows up to 45% reduction in L2 instruction-miss rate (19.6% on average), resulting in up to 8% speedup (4.7% on average). These gains are up to 4X larger than full-hardware record-and-replay prefetchers, while needing two orders of magnitude smaller on-chip storage.

cs.PF↗

arXiv2025

Dynamics of discrete random Burgers-Huxley systems: attractor convergence and finite-dimensional approximations

In this paper, we apply the implicit Euler scheme to discretize the (random) Burgers-Huxley equation and prove the existence of a numerical attractor for the discrete Burgers-Huxley system. We establish upper semi-convergence of the numerical attractor to the global attractor as the step size tends to zero. We also provide finite-dimensional approximations for the three attractors (global, numerical and random) and prove the existence of truncated attractors as the state space dimension goes to infinity. Finally, we prove the existence of a random attractor and establish that the truncated random attractor upper semi-converges to the truncated global attractor as the noise intensity tends to zero.

math.AP↗

arXiv2019

Low-Rank Deep Convolutional Neural Network for Multi-Task Learning

In this paper, we propose a novel multi-task learning method based on the deep convolutional network. The proposed deep network has four convolutional layers, three max-pooling layers, and two parallel fully connected layers. To adjust the deep network to multi-task learning problem, we propose to learn a low-rank deep network so that the relation among different tasks can be explored. We proposed to minimize the number of independent parameter rows of one fully connected layer to explore the relations among different tasks, which is measured by the nuclear norm of the parameter of one fully connected layer, and seek a low-rank parameter matrix. Meanwhile, we also propose to regularize another fully connected layer by sparsity penalty, so that the useful features learned by the lower layers can be selected. The learning problem is solved by an iterative algorithm based on gradient descent and back-propagation algorithms. The proposed algorithm is evaluated over benchmark data sets of multiple face attribute prediction, multi-task natural language processing, and joint economics index predictions. The evaluation results show the advantage of the low-rank deep CNN model over multi-task problems.

cs.LG↗

arXiv2018

Self-consistency and covariance of light-front quark models: testing via $P$, $V$ and $A$ meson decay constants, and $P\to P$ weak transition form factors

In this paper, we test the self-consistencies of the standard and the covariant light-front quark model and study the zero-mode issue via the decay constants of pseudoscalar ($P$), vector ($V$) and axial-vector ($A$) mesons, as well as the $P\to P$ weak transition form factors. With the traditional type-I correspondence between the manifestly covariant and the light-front approach, the resulting $f_{V}$ as well as $f_{^1\!A}$ and $f_{^3\!A}$ obtained with the $\lbd=0$ and $\lbd=\pm$ polarization states are different from each other, which presents a challenge to the self-consistency of the covariant light-front quark model. However, such a self-consistency problem can be "resolved" within the type-II scheme, which requires an additional replacement $M\to M_0$ relative to the type-I case. Moreover, the replacement $M\to M_0$ is also essential for the self-consistency of the standard light-front quark model. In the type-II scheme, the valence contributions to the physical quantities~(${\cal Q}$) considered in this paper are alway the same as that obtained in the standard light-front quark model, $[{\cal Q}]_{\rm val.}=[{\cal Q}]_{\rm SLF}$, and the zero-mode contributions to $f_{V,^1\!A,^3\!A}$ and $f_-(q^2)$ exist only formally but vanish numerically, which implies further that $[{\cal Q}]_{\rm val.}\dot{=} [{\cal Q}]_{\rm full}$. In addition, the manifest covariance of the covariant light-front quark model is violated in the traditional type-I scheme, but can be recovered by taking the type-II scheme.

hep-ph↗

arXiv2018

Cross-domain attribute representation based on convolutional neural network

In the problem of domain transfer learning, we learn a model for the predic-tion in a target domain from the data of both some source domains and the target domain, where the target domain is in lack of labels while the source domain has sufficient labels. Besides the instances of the data, recently the attributes of data shared across domains are also explored and proven to be very helpful to leverage the information of different domains. In this paper, we propose a novel learning framework for domain-transfer learning based on both instances and attributes. We proposed to embed the attributes of dif-ferent domains by a shared convolutional neural network (CNN), learn a domain-independent CNN model to represent the information shared by dif-ferent domains by matching across domains, and a domain-specific CNN model to represent the information of each domain. The concatenation of the three CNN model outputs is used to predict the class label. An iterative algo-rithm based on gradient descent method is developed to learn the parameters of the model. The experiments over benchmark datasets show the advantage of the proposed model.

cs.CV↗

arXiv2018

Decay constants of pseudoscalar and vector mesons with improved holographic wavefunction

We calculate the decay constants of light and heavy-light pseudoscalar and vector mesons with improved soft-wall holographic wavefuntions, which take into account the effects of both quark masses and dynamical spins. We find that the predicted decay constants, especially for the ratio $f_V/f_P$, based on light-front holographic QCD, can be significantly improved, once the dynamical spin effects are taken into account by introducing the helicity-dependent wavefunctions. We also perform detailed $χ^2$ analyses for the holographic parameters ({\it i.e.} the mass-scale parameter $κ$ and the quark masses), by confronting our predictions with the data for the charged-meson decay constants and the meson spectra. The fitted values for these parameters are generally in agreement with those obtained by fitting to the Regge trajectories. At the same time, most of our results for the decay constants and their ratios agree with the data as well as the predictions based on lattice QCD and QCD sum rule approaches, with only a few exceptions observed.

hep-ph↗

arXiv2018

Domain transfer convolutional attribute embedding

In this paper, we study the problem of transfer learning with the attribute data. In the transfer learning problem, we want to leverage the data of the auxiliary and the target domains to build an effective model for the classification problem in the target domain. Meanwhile, the attributes are naturally stable cross different domains. This strongly motives us to learn effective domain transfer attribute representations. To this end, we proposed to embed the attributes of the data to a common space by using the powerful convolutional neural network (CNN) model. The convolutional representations of the data points are mapped to the corresponding attributes so that they can be effective embedding of the attributes. We also represent the data of different domains by a domain-independent CNN, ant a domain-specific CNN, and combine their outputs with the attribute embedding to build the classification model. An joint learning framework is constructed to minimize the classification errors, the attribute mapping error, the mismatching of the domain-independent representations cross different domains, and to encourage the the neighborhood smoothness of representations in the target domain. The minimization problem is solved by an iterative algorithm based on gradient descent. Experiments over benchmark data sets of person re-identification, bankruptcy prediction, and spam email detection, show the effectiveness of the proposed method.

cs.LG↗

arXiv2017

$K^0-\bar{K}^0$ mixing in the minimal flavor-violating two-Higgs-doublet models

The two-Higgs-doublet model (2HDM), as one of the simplest extensions of the Standard Model (SM), is obtained by adding another scalar doublet to the SM, and is featured by a pair of charged scalars, which could affect many low-energy processes. In the "Higgs basis" for a generic 2HDM, only one scalar doublet gets a nonzero vacuum expectation value and, under the criterion of minimal flavor violation, the other one is fixed to be either color-singlet or color-octet, which are named as the type-III and the type-C 2HDM, respectively. In this paper, we study the charged-scalar effects of these two models on the $K^0-\bar{K}^0$ mixing, an ideal process to probe New Physics (NP) beyond the SM. Firstly, we perform a complete one-loop computation of the box diagrams relevant to the $K^0-\bar{K}^0$ mixing, keeping the mass and momentum of the external strange quark up to the second order. Together with the up-to-date theoretical inputs, we then give a detailed phenomenological analysis, in the cases of both real and complex Yukawa couplings of the charged scalars to quarks. The parameter spaces allowed by the current experimental data on the mass difference $Δm_K$ and the CP-violating parameter $ε_K$ are obtained and the differences between these two 2HDMs are investigated, which are helpful to distinguish them from each other from a phenomenological point of view.

hep-ph↗

arXiv2012

$B_s^0-\bar B_s^0$ mixing in a family non-universal $Z^{\prime}$ model revisited

Motivated by the very recent measurements performed at the LHCb and the Tevatron of the $B_s^0-\bar B_s^0$ mixing, in this paper we revisit it in a family non-universal $Z^{\prime}$ model, to check if a simultaneous explanation for all the mixing observables, especially for the like-sign dimuon charge asymmetry observed by the D0 collaboration, could be made in such a specific model. In the first scenario where the $Z^\prime$ boson contributes only to the off-diagonal element $M_{12}^s$, it is found that, once the combined constraints from $ΔM_s$, $ϕ_s$ and $ΔΓ_s$ are imposed, the model could not explain the measured flavour-specific CP asymmetry $a_{fs}^s$, at least within its $1σ$ ranges. In the second scenario where the NP contributes also to the absorptive part $Γ_{12}^s$ via tree-level $Z^\prime$-induced $b\to c\bar{c}s$ operators, we find that, with the constraints from $ΔM_s$, $ϕ_s$ and the indirect CP asymmetry in $\bar{B}_d\to J/ψK_S$ taken into account, the present measured $1σ$ experimental ranges for $a_{fs}^s$ could not be reproduced too. Thus, such a specific $Z^\prime$ model with our specific assumptions could not simultaneously reconcile all the present data on $B_s^0-\bar B_s^0$ mixing. Future improved measurements from the LHCb and the proposed superB experiments, especially of the flavour-specific CP asymmetries, are expected to shed light on the issue.

hep-ph↗

arXiv2012

Charmless $B_s\to PP, PV, VV$ Decays Based on the six-quark Effective Hamiltonian with Strong Phase Effects II

We provide a systematic study of charmless $B_s \to PP, PV, VV$ decays ($P$ and $V$ denote pseudoscalar and vector mesons, respectively) based on an approximate six-quark operator effective Hamiltonian from QCD. The calculation of the relevant hard-scattering kernels is carried out, the resulting transition form factors are consistent with the results of QCD sum rule calculations. By taking into account important classes of power corrections involving "chirally-enhanced" terms and the vertex corrections as well as weak annihilation contributions with non-trivial strong phase, we present predictions for the branching ratios and CP asymmetries of $B_s$ decays into PP, PV and VV final states, and also for the corresponding polarization observables in VV final states. It is found that the weak annihilation contributions with non-trivial strong phase have remarkable effects on the observables in the color-suppressed and penguin-dominated decay modes. In addition, we discuss the SU(3) flavor symmetry and show that the symmetry relations are generally respected.

hep-ph↗

arXiv2011

Determination of the strong coupling $g_{B^* Bπ}$ from semi-leptonic $B\to π\ell ν$ decay

According to heavy-meson chiral perturbation theory, the vector form factor $f_+(q^2)$ of exclusive semi-leptonic decay $B\to π\ell ν$ is closely related, at least in the soft-pion region (i.e., $q^{2} \sim (m_B-m_π)^2$), to the strong coupling $g_{B^* Bπ}$ or the normalized coupling $\hat g$. Combining the precisely measured $q^2$ spectrum of $B\to π\ell ν$ decay by the BaBar and Belle collaborations with several parametrizations of the form factor $f_+(q^2)$, we can extract these couplings from the residue of the form factor at the $B^*$ pole, which relies on an extrapolation of the form factor from the semi-leptonic region to the unphysical point $q^2=m_{B^*}^2$. Comparing the extracted values with the other experimental and theoretical estimates, we can test these various form-factor parametrizations, which differ from each other by the amount of physical information embedded in. It is found that the extracted values based on the BK, BZ and BCL parametrizations are consistent with each other and roughly in agreement with the other theoretical and lattice estimates, while the BGL ansatz, featured by a spurious, unwanted pole at the threshold of the cut, gives a neatly larger value.

hep-ph↗

arXiv2010

Charmless $B\to PP, PV, VV$ Decays Based on the six-quark Effective Hamiltonian with Strong Phase Effects

Based on an approximate six-quark operator effective Hamiltonian from perturbative QCD, we present a systematical study of charmless $B \to PP, PV, VV$ decays ($P$ and $V$ denoting pseudoscalar and vector mesons, respectively). The calculation of the relevant hard-scattering kernels is completed, the resulting transition form factors are consistent with the results of QCD sum rule calculation. Important classes of power corrections include "chirally-enhanced" terms, vertex corrections and weak annihilation contributions with non-trivial strong phase. With these considerations, predictions are presented for the branching ratios and CP asymmetries of B-meson decays into PP, PV and VV final states, and also for the corresponding polarization observables in VV final states. Several decay modes and observables, which are of particular interest phenomenologically, are discussed in detail, including the effect of annihilation amplitude with strong phase, the $ππ$, $πK$ and $πρ$ systems, the longitudinal polarization fraction $f_L$ in $ρK^*$ and $ϕK^*$ systems and so on. It is observed that predictions in our framework generally agree with the current experimental data.

hep-ph↗

arXiv2010

Constraints on B--->pi,K transition form factors from exclusive semileptonic D-meson decays

According to the heavy-quark flavour symmetry, the $B\to π, K$ transition form factors could be related to the corresponding ones of D-meson decays near the zero recoil point. With the recent precisely measured exclusive semileptonic decays $D \to π\ell ν$ and $D\to K \ell ν$, we perform a phenomenological study of $B \to π, K$ transition form factors based on this symmetry. Using BK, BZ and Series Expansion parameterizations of the form factor slope, we extrapolate $B \to π, K$ transition form factors from $q^{2}_{max}$ to $q^{2}=0$. It is found that, although being consistent with each other within error bars, the central values of our results for $B \to π, K$ form factors at $q^2=0$, $f_+^{B\to π, K}(0)$, are much smaller than predictions of the QCD light-cone sum rules, but are in good agreements with the ones extracted from hadronic B-meson decays within the SCET framework. Moreover, smaller form factors are also favored by the QCD factorization approach for hadronic B-meson decays.

hep-ph↗

arXiv2008

QCD Factorization Based on Six-Quark Operator Effective Hamiltonian from Perturbative QCD and Charmless Bottom Meson Decays $B_{(s)}\to ππ,πK, KK$

The charmless bottom meson decays are systematically investigated based on an approximate six quark operator effective Hamiltonian from perturbative QCD. It is shown that within this framework the naive QCD factorization method provides a simple way to evaluate the hadronic matrix elements of two body mesonic decays. The singularities caused by on mass-shell quark propagator and gluon exchanging interaction are appropriately treated. Such a simple framework allows us to make theoretical predictions for the decay amplitudes with reasonable input parameters. The resulting theoretical predictions for all the branching ratios and CP asymmetries in the charmless $B^0, B^+, B_s\to ππ, πK, KK$ decays are found to be consistent with the current experimental data except for a few decay modes. The observed large branching ratio in $B\to π^0π^0$ decay remains a puzzle though the predicted branching ratio may be significantly improved by considering the large vertex corrections in the effective Wilson coefficients. More precise measurements of charmless bottom meson decays, especially on CP-violations in $B\to K K$ and $B_s\to ππ, πK, KK$ decay modes, will provide a useful test and guide us to a better understanding on perturbative and nonperturbative QCD.

hep-ph↗