arXiv subjects

Ji Xu

Publications and source records attributed to Ji Xu.

At least 91 records · Page 5Linked to original sources

arXiv2019

Non-iterative Label Propagation in Optimal Leading Forest

Graph based semi-supervised learning (GSSL) has intuitive representation and can be improved by exploiting the matrix calculation. However, it has to perform iterative optimization to achieve a preset objective, which usually leads to low efficiency. Another inconvenience lying in GSSL is that when new data come, the graph construction and the optimization have to be conducted all over again. We propose a sound assumption, arguing that: the neighboring data points are not in peer-to-peer relation, but in a partial-ordered relation induced by the local density and distance between the data; and the label of a center can be regarded as the contribution of its followers. Starting from the assumption, we develop a highly efficient non-iterative label propagation algorithm based on a novel data structure named as optimal leading forest (LaPOLeaF). The major weaknesses of the traditional GSSL are addressed by this study. We further scale LaPOLeaF to accommodate big data by utilizing block distance matrix technique, parallel computing, and Locality-Sensitive Hashing (LSH). Experiments on large datasets have shown the promising results of the proposed methods.

cs.LG↗

arXiv2018

Benefits of over-parameterization with EM

Expectation Maximization (EM) is among the most popular algorithms for maximum likelihood estimation, but it is generally only guaranteed to find its stationary points of the log-likelihood objective. The goal of this article is to present theoretical and empirical evidence that over-parameterization can help EM avoid spurious local optima in the log-likelihood. We consider the problem of estimating the mean vectors of a Gaussian mixture model in a scenario where the mixing weights are known. Our study shows that the global behavior of EM, when one uses an over-parameterized model in which the mixing weights are treated as unknown, is better than that when one uses the (correct) model with the mixing weights fixed to the known values. For symmetric Gaussians mixtures with two components, we prove that introducing the (statistically redundant) weight parameters enables EM to find the global maximizer of the log-likelihood starting from almost any initial mean parameters, whereas EM without this over-parameterization may very often fail. For other Gaussian mixtures, we provide empirical evidence that shows similar behavior. Our results corroborate the value of over-parameterization in solving non-convex optimization problems, previously observed in other domains.

cs.LG↗

arXiv2018

Matching the Quasi Meson Distribution Amplitude in RI/MOM scheme

The $x$-dependence of light-cone distribution amplitude (LCDA) can be directly calculated from a quasi distribution amplitude (DA) in lattice QCD within the framework of large-momentum effective theory (LaMET). In this paper, we study the one-loop renormalization of the quasi-DA in the regularization-independent momentum subtraction (RI/MOM) scheme. The renormalization factor for the quasi parton distribution function can be used to renormalize the quasi-DA provided that they are implemented on lattice and in perturbation theory in the same manner. We derive the one-loop matching coefficient that matches quasi-DA in the RI/MOM scheme onto LCDA in the $\overline{\rm MS}$ scheme. Our result provides the crucial step to extract the LCDAs from lattice matrix elements of quasi-DAs.

hep-ph↗

arXiv2018

Approximate Message Passing for Amplitude Based Optimization

We consider an $\ell_2$-regularized non-convex optimization problem for recovering signals from their noisy phaseless observations. We design and study the performance of a message passing algorithm that aims to solve this optimization problem. We consider the asymptotic setting $m,n \rightarrow \infty$, $m/n \rightarrow δ$ and obtain sharp performance bounds, where $m$ is the number of measurements and $n$ is the signal dimension. We show that for complex signals the algorithm can perform accurate recovery with only $m=\left ( \frac{64}{π^2}-4\right)n\approx 2.5n$ measurements. Also, we provide sharp analysis on the sensitivity of the algorithm to noise. We highlight the following facts about our message passing algorithm: (i) Adding $\ell_2$ regularization to the non-convex loss function can be beneficial even in the noiseless setting; (ii) spectral initialization has marginal impact on the performance of the algorithm.

cs.IT↗

arXiv2018

Light-cone distribution amplitudes of vector meson in large momentum effective theory

We investigate the leading twist light-cone distribution amplitudes (LCDAs) of vector meson in the framework of large momentum effective theory. We derive the matching equation for the LCDAs and quasi distribution amplitudes. The matching coefficients are determined to one loop accuracy, both in the ultraviolet cut-off and dimensional regularization schemes. This calculation provides the possibility of studying the full $x$ behavior of LCDAs and extracting LCDAs of vector mesons from lattice simulations.

hep-ph↗

arXiv2018

Weak Decays of Triply Heavy Baryons

After the experimental establishment of doubly heavy baryons, baryons with three quarks are the last missing pieces of the lowest-lying baryon multiplets in quark model. In this work we study semileptonic and nonleptonic weak decays of triply heavy baryons, $Ω_{ccc}^{++}, Ω_{ccb}^{+}, Ω_{cbb}^{0}, Ω_{bbb}^{-}$. Decay amplitudes for various channels are parametrized in terms of a few SU(3) irreducible amplitudes. We point out that branching fractions for Cabibbo allowed processes, $Ω_{ccc}\to (Ξ_{cc}^{++} \overline K^0, Ξ_{cc}^{++}K^-π^+, Ω_{cc}^{+}π^+, Ξ_{c}^+ D^+, Ξ_{c}^{\prime} D^+, Λ_c D^+\overline K^0, Ξ_{c}^+ D^0 π^+, Ξ_{c}^0 D^+π^+)$ may reach a few percents. We suggest our experimental colleagues to perform a search at hadron colliders and the electron and positron collisions in future, which will presumably lead to discoveries of triply heavy baryons and complete the baryon multiplets. Using the expanded amplitudes, we derive a number of relations for the partial widths which can be examined in future.

hep-ph↗

arXiv2018

Optimization-based AMP for Phase Retrieval: The Impact of Initialization and $\ell_2$-regularization

We consider an $\ell_2$-regularized non-convex optimization problem for recovering signals from their noisy phaseless observations. We design and study the performance of a message passing algorithm that aims to solve this optimization problem. We consider the asymptotic setting $m,n \rightarrow \infty$, $m/n \rightarrow δ$ and obtain sharp performance bounds, where $m$ is the number of measurements and $n$ is the signal dimension. We show that for complex signals the algorithm can perform accurate recovery with only $m= \left(\frac{64}{π^2}-4\right)n \approx 2.5n$ measurements. Also, we provide sharp analysis on the sensitivity of the algorithm to noise. We highlight the following facts about our message passing algorithm: (i) Adding $\ell_2$ regularization to the non-convex loss function can be beneficial. (ii) Spectral initialization has marginal impact on the performance of the algorithm. The sharp analyses in this paper, not only enable us to compare the performance of our method with other phase recovery schemes, but also shed light on designing better iterative algorithms for other non-convex optimization problems.

cs.IT↗

arXiv2017

Weak Decays of Doubly Heavy Baryons: Multi-body Decay Channels

The newly-discovered $Ξ_{cc}^{++}$ decays into the $ Λ_{c}^+ K^-π^+π^+$, but the experimental data has indicated that this decay is not saturated by any two-body intermediate state. In this work, we analyze the multi-body weak decays of doubly heavy baryons $Ξ_{cc}$, $Ω_{cc}$, $Ξ_{bc}$, $Ω_{bc}$, $Ξ_{bb}$ and $Ω_{bb}$, in particular the three-body nonleptonic decays and four-body semileptonic decays. We classify various decay modes according to the quark-level transitions and present an estimate of the typical branching fractions for a few golden decay channels. Decay amplitudes are then parametrized in terms of a few SU(3) irreducible amplitudes. With these amplitudes, we find a number of relations for decay widths, which can be examined in future.

hep-ph↗

arXiv2017

Weak Decays of Doubly Heavy Baryons: SU(3) Analysis

Motivated by the recent LHCb observation of doubly-charmed baryon $Ξ_{cc}^{++}$ in the $Λ_c^+ K^-π^+π^+$ final state, we analyze the weak decays of doubly heavy baryons $Ξ_{cc}$, $Ω_{cc}$, $Ξ_{bc}$, $Ω_{bc}$, $Ξ_{bb}$ and $Ω_{bb}$ under the flavor SU(3) symmetry. Decay amplitudes for various semileptonic and nonleptonic decays are parametrized in terms of a few SU(3) irreducible amplitudes. We find a number of relations or sum rules between decay widths and CP asymmetries, which can be examined in future measurements at experimental facilities like LHC, Belle II and CEPC. Moreover once a few decay branching fractions are measured in future, some of these relations may provide hints for exploration of new decay modes.

hep-ph↗

arXiv2017

An Improved Residual LSTM Architecture for Acoustic Modeling

Long Short-Term Memory (LSTM) is the primary recurrent neural networks architecture for acoustic modeling in automatic speech recognition systems. Residual learning is an efficient method to help neural networks converge easier and faster. In this paper, we propose several types of residual LSTM methods for our acoustic modeling. Our experiments indicate that, compared with classic LSTM, our architecture shows more than 8% relative reduction in Phone Error Rate (PER) on TIMIT tasks. At the same time, our residual fast LSTM approach shows 4% relative reduction in PER on the same task. Besides, we find that all this architecture could have good results on THCHS-30, Librispeech and Switchboard corpora.

cs.CL↗

arXiv2017

Cumulants of event-by-event net-strangeness distributions in Au+Au collisions at $\sqrt{s_\mathrm{NN}}$=7.7-200 GeV from UrQMD model

Fluctuations of conserved quantities, such as baryon, electric charge and strangeness number, are sensitive observables in heavy-ion collisions to search for the QCD phase transition and critical point. In this paper, we performed a systematical analysis on the various cumulants and cumulant ratios of event-by-event net-strangeness distributions in Au+Au collisions at $\sqrt{s_{NN}}$=7.7, 11.5, 19.6, 27, 39, 62.4 and 200 GeV from UrQMD model. We performed a systematical study on the contributions from various strange baryons and mesons to the net-strangeness fluctuations. The results demonstrate that the cumulants and cumulant ratios of net-strangeness distributions extracted from different strange particles show very different centrality and energy dependence behavior. By comparing with the net-kaon fluctuations, we found that the strange baryons play an important role in the fluctuations of net-strangeness. This study can provide useful baselines to study the QCD phase transition and search for the QCD critical point by using the fluctuations of net-strangeness in heavy-ion collisions experiment. It can help us to understand non-critical physics contributions to the fluctuations of net-strangeness.

nucl-ex↗

arXiv2017

Relativistic corrections to light-cone distribution amplitudes of S-wave $B_c$ mesons and heavy quarkonia

In collinear factorization, light-cone distribution amplitudes (LCDAs) are key ingredients to calculate the production rate of a hadron in high energy exclusive processes. For a doubly-heavy meson system (such as $B_c, J/ψ, Υ$ etc), the LCDAs contain perturbative scales that can be integrated out and then are re-factorized into products of perturbatively calculable distribution parts and non-relativistic QCD matrix elements. In this re-factorization scheme, the LCDAs are known at next-to-leading order in the strong coupling constant $α_s$ and at leading order in the velocity expansion. In this work, we calculate the ${\cal O}( { v}^2)$ corrections to twist-2 LCDAs of S-wave $B_c$ mesons. These results are applicable to heavy quarkonia like $η_{c,b}$, $J/ψ$ and $Υ$ by setting $m_b=m_c$. We apply these relativistically corrected LCDAs to study their inverse moments and a few Gegenbauer moments which are important for phenomenological study. We point out that the relativistic corrections are sizable, and comparable with the next-to-leading order radiative corrections. These results for LCDAs are useful in future theoretical analyses of the productions of heavy quarkonia and $B_c$ mesons.

hep-ph↗

arXiv2016

Boson-Fermion Duality from Discontinuous Gauge Symmetry

By extending local U(1) gauge symmetry to discontinuous case, we find that under one special discontinuous U(1) gauge transformation the symmetric and antisymmetric wave functions can transform into each other in one dimensional quantum mechanics. The free spinless fermionic system and bosonic system with $δ$-type vector gauge potential are proved to be equivalent. The relation also holds in higher space-time dimensions.

hep-th↗

arXiv2016

Energy Dependence of Moments of Net-Proton, Net-Kaon, and Net-Charge Multiplicity Distributions at STAR

One of the main goals of the RHIC Beam Energy Scan (BES) program is to study the QCD phase structure, which includes the search for the QCD critical point, over a wide range of chemical potential. Theoretical calculations predict that fluctuations of conserved quantities, such as baryon number (B), charge (Q), and strangeness (S), are sensitive to the correlation length of the dynamical system. Experimentally, higher moments of multiplicity distributions have been utilized to search for the QCD critical point in heavy-ion collisions. In this paper, we report recent efficiency-corrected cumulants and cumulants ratios of the net- proton, net-kaon, and net-charge multiplicity distributions in Au+Au collisions at 7.7, 11.5, 14.5, 19.6, 27, 39, 62.4, and 200 GeV collected in the years 2010, 2011, and 2014 with STAR at RHIC. The centrality and energy dependence of the cumulants up to the fourth order, as well as their ratios, are presented. Furthermore, the comparisons with baseline calculations (Poisson) and non-critical-point models (UrQMD) will also be discussed.

hep-ex↗

arXiv2016

Higher Moments of Net-Kaon Multiplicity Distributions at STAR

Fluctuations of conserved quantities such as baryon number (B), electric charge number (Q), and strangeness number (S), are sensitive to the correlation length and can be used to probe non-gaussian fluctuations near the critical point. Experimentally, higher moments of the multiplicity distributions have been used to search for the QCD critical point in heavy-ion collisions. In this paper, we report the efficiency-corrected cumulants and their ratios of mid- rapidity (|y| < 0.5) net-kaon multiplicity distributions in Au+Au collisions at 7.7, 11.5, 14.5, 19.6, 27, 39, 62.4, and 200 GeV collected in 2010, 2011, and 2014 with STAR at RHIC. The centrality and energy dependence of the cumulants and their ratios, are presented. Furthermore, the comparisons with baseline calculations (Poisson) and non-critical-point models (UrQMD) are also discussed.

hep-ex↗

arXiv2016

Global analysis of Expectation Maximization for mixtures of two Gaussians

Expectation Maximization (EM) is among the most popular algorithms for estimating parameters of statistical models. However, EM, which is an iterative algorithm based on the maximum likelihood principle, is generally only guaranteed to find stationary points of the likelihood objective, and these points may be far from any maximizer. This article addresses this disconnect between the statistical principles behind EM and its algorithmic properties. Specifically, it provides a global analysis of EM for specific models in which the observations comprise an i.i.d. sample from a mixture of two Gaussians. This is achieved by (i) studying the sequence of parameters from idealized execution of EM in the infinite sample limit, and fully characterizing the limit points of the sequence in terms of the initial parameters; and then (ii) based on this convergence analysis, establishing statistical consistency (or lack thereof) for the actual sequence of parameters produced by EM.

math.ST↗

arXiv2016

The leading twist light-cone distribution amplitudes for the S-wave and P-wave Bc mesons

The light-cone distribution amplitudes (LCDAs) serve as important non-perturbative inputs for the study of hard exclusive processes. In this paper, we calculate ten LCDAs at twist-2 for the S-wave and P-wave $B_c$ mesons up to the next-to-leading order (NLO) of the strong coupling $α_s$ and leading order of the velocity expansion. Each one of these ten LCDAs is expressed as a product of a perturbatively calculable distribution and a universal NRQCD matrix-element. By use of the spin symmetry, only two NRQCD matrix-elements will be involved. The reduction of the number of non-perturbative inputs will improve the predictive power of collinear factorization.

hep-ph↗

arXiv2016

Cumulants of Net-Proton, Net-Kaon and Net-Charge Multiplicity Distributions in Au+Au Collisions at RHIC BES Energies from UrQMD Model

Fluctuations of conserved quantities are sensitive observables to probe the signature of QCD phase transition and critical point in heavy-ion collisions. With the UrQMD model, we have studied the centrality and energy dependence of various order cumulants and cumulant ratios (up to fourth order) of net-proton,net-charge and net-kaon multiplicity distributions in Au+Au collisions at $\sqrt{s_{NN}}$= 7.7, 11.5, 19.6, 27, 39, 62.4, 200 GeV. The model results show that the production mechanism of the particles and anti-particles have significant impacts on the cumulants of net-particles multiplicity distributions and show strong energy dependence. We also made comparisons between model calculations and experimental data measured in the first phase of the beam energy scan (BES) program by the STAR experiment at RHIC. The comparisons indicate that the baryon conservation effect strongly suppress the cumulants of net-proton distributions at low energies and the non-monotonic energy dependence for the net-proton {\KV} at the most central Au+Au collisions measured by the STAR experiment can not be described by the UrQMD model. Since there has no physics of QCD phase transition and QCD critical point implemented in the UrQMD, the model results provide us baselines and qualitative estimates about the non-critical background contributions to the fluctuations observables in heavy-ion collisions.

nucl-ex↗