arXiv subjects

Hao Liang

Publications and source records attributed to Hao Liang.

At least 163 records · Page 9Linked to original sources

arXiv2023

ParticleNet and its application on CEPC Jet Flavor Tagging

Identification of quark flavor is essential for collider experiments in high-energy physics, relying on the flavor tagging algorithm. In this study, using a full simulation of the Circular Electron Positron Collider (CEPC), we investigated the flavor tagging performance of two different algorithms: ParticleNet, originally developed at CMS, and LCFIPlus, the current flavor tagging algorithm employed at CEPC. Compared to LCFIPlus, ParticleNet significantly enhances flavor tagging performance, resulting in a significant improvement in benchmark measurement accuracy, i.e., a 36% improvement for $ν\barνH\to c\bar{c}$ measurement and a 75% improvement for $|V_{cb}|$ measurement via W boson decay when CEPC operates as a Higgs factory at the center-of-mass energy of 240 GeV and integrated luminosity of 5.6 $ab^{-1}$. We compared the performance of ParticleNet and LCFIPlus at different vertex detector configurations, observing that the inner radius is the most sensitive parameter, followed by material budget and spatial resolution.

hep-ex↗

arXiv2023

Visualizing chest X-ray dataset biases using GANs

Recent work demonstrates that images from various chest X-ray datasets contain visual features that are strongly correlated with protected demographic attributes like race and gender. This finding raises issues of fairness, since some of these factors may be used by downstream algorithms for clinical predictions. In this work, we propose a framework, using generative adversarial networks (GANs), to visualize what features are most different between X-rays belonging to two demographic subgroups.

eess.IV↗

arXiv2023

Benchmarking Algorithmic Bias in Face Recognition: An Experimental Approach Using Synthetic Faces and Human Evaluation

We propose an experimental method for measuring bias in face recognition systems. Existing methods to measure bias depend on benchmark datasets that are collected in the wild and annotated for protected (e.g., race, gender) and non-protected (e.g., pose, lighting) attributes. Such observational datasets only permit correlational conclusions, e.g., "Algorithm A's accuracy is different on female and male faces in dataset X.". By contrast, experimental methods manipulate attributes individually and thus permit causal conclusions, e.g., "Algorithm A's accuracy is affected by gender and skin color." Our method is based on generating synthetic faces using a neural face generator, where each attribute of interest is modified independently while leaving all other attributes constant. Human observers crucially provide the ground truth on perceptual identity similarity between synthetic image pairs. We validate our method quantitatively by evaluating race and gender biases of three research-grade face recognition models. Our synthetic pipeline reveals that for these algorithms, accuracy is lower for Black and East Asian population subgroups. Our method can also quantify how perceptual changes in attributes affect face identity distances reported by these models. Our large synthetic dataset, consisting of 48,000 synthetic face image pairs (10,200 unique synthetic faces) and 555,000 human annotations (individual attributes and pairwise identity comparisons) is available to researchers in this important area.

cs.CV↗

arXiv2023

Learning an Interpretable End-to-End Network for Real-Time Acoustic Beamforming

Recently, many forms of audio industrial applications, such as sound monitoring and source localization, have begun exploiting smart multi-modal devices equipped with a microphone array. Regrettably, model-based methods are often difficult to employ for such devices due to their high computational complexity, as well as the difficulty of appropriately selecting the user-determined parameters. As an alternative, one may use deep network-based methods, but these are often difficult to generalize, nor can they generate the desired beamforming map directly. In this paper, a computationally efficient acoustic beamforming algorithm is proposed, which may be unrolled to form a model-based deep learning network for real-time imaging, here termed the DAMAS-FISTA-Net. By exploiting the natural structure of an acoustic beamformer, the proposed network inherits the physical knowledge of the acoustic system, and thus learns the underlying physical properties of the propagation. As a result, all the network parameters may be learned end-to-end, guided by a model-based prior using back-propagation. Notably, the proposed network enables an excellent interpretability and the ability of being able to process the raw data directly. Extensive numerical experiments using both simulated and real-world data illustrate the preferable performance of the DAMAS-FISTA-Net as compared to alternative approaches.

cs.SD↗

arXiv2023

A Distribution Optimization Framework for Confidence Bounds of Risk Measures

We present a distribution optimization framework that significantly improves confidence bounds for various risk measures compared to previous methods. Our framework encompasses popular risk measures such as the entropic risk measure, conditional value at risk (CVaR), spectral risk measure, distortion risk measure, equivalent certainty, and rank-dependent expected utility, which are well established in risk-sensitive decision-making literature. To achieve this, we introduce two estimation schemes based on concentration bounds derived from the empirical distribution, specifically using either the Wasserstein distance or the supremum distance. Unlike traditional approaches that add or subtract a confidence radius from the empirical risk measures, our proposed schemes evaluate a specific transformation of the empirical distribution based on the distance. Consequently, our confidence bounds consistently yield tighter results compared to previous methods. We further verify the efficacy of the proposed framework by providing tighter problem-dependent regret bound for the CVaR bandit.

cs.LG↗

arXiv2023

Regret Bounds for Risk-sensitive Reinforcement Learning with Lipschitz Dynamic Risk Measures

We study finite episodic Markov decision processes incorporating dynamic risk measures to capture risk sensitivity. To this end, we present two model-based algorithms applied to \emph{Lipschitz} dynamic risk measures, a wide range of risk measures that subsumes spectral risk measure, optimized certainty equivalent, distortion risk measures among others. We establish both regret upper bounds and lower bounds. Notably, our upper bounds demonstrate optimal dependencies on the number of actions and episodes, while reflecting the inherent trade-off between risk sensitivity and sample complexity. Additionally, we substantiate our theoretical results through numerical experiments.

cs.LG↗

arXiv2023

The non-Archimedean Nirgendsnegativsemidefinitheitsstellensatz is not true

Klep and Schweighofer asked whether the Nirgendsnegativsemide-finitheitsstellensatz holds for a symmetric noncommutative polynomial whose evaluations at bounded self-adjoint operators on any nontrivial Hilbert space are not negative semidefinite. We provide an example to show the open problem has a negative answer.

math.OA↗

arXiv2023

X-ray Recognition: Patient identification from X-rays using a contrastive objective

Recent research demonstrates that deep learning models are capable of precisely extracting bio-information (e.g. race, gender and age) from patients' Chest X-Rays (CXRs). In this paper, we further show that deep learning models are also surprisingly accurate at recognition, i.e., distinguishing CXRs belonging to the same patient from those belonging to different patients. These findings suggest potential privacy considerations that the medical imaging community should consider with the proliferation of large public CXR databases.

eess.IV↗

arXiv2023

Laser-assisted Fano resonance: attosecond quantum control and dynamical imaging

A Fano resonance arises from the pathway interference between discrete and continuum states, playing a fundamental role in many branches of physics, chemistry and material science. Here, we introduce the concept of a laser-assisted Fano resonance, created from two interferometric pathways that are coupled together by an additional laser field, which introduces a controllable phase delay between them and results in a generalized Fano lineshape that can be actively controlled on the {\it attosecond} time scale. Based on our experimental results of unprecedented resolution, we dynamically image a resonant electron wave packet during its evolution directly in the time domain, extracting both the amplitude and the phase, which allows for the measurement of the {\it resonant} photoionization time delay. Ab-initio calculations and simulations employing a physically transparent two-level model agree with our experimental results, laying the groundwork for extending our concepts into attosecond quantum control of complex systems.

quant-ph↗

arXiv2022

Optimization of tracker configuration for the CEPC

We investigate the tracker configuration optimization for the Circular Electron Position Collider (CEPC), a proposed Higgs and $Z$ factory. Fixing the construction cost comparable to that of the baseline detector design and considering the benchmark channels ($Z\rightarrow f\bar{f}$, $WW$ fusion with $H\rightarrow f\bar{f}$, $ZH\rightarrowν\barνf\bar{f}$, and $t\bar{t}\rightarrow b\bar{b}μν_μud$) of various operating modes of the CEPC, we obtain the optimal tracker radius that provides the best average resolution of the track momentum or jet energy. The optimal tracker radii for track momentum resolution range from 1.59\,m to 1.73\,m and for jet energy resolution from 1.82\,m to 1.97\,m, depending on the benchmark channels. Compared to the jets, the tracks prefer a smaller radius and a longer length because the track momentum resolution degrades more significantly than jet energy resolution in the forward region. The benchmark channel for $Z$-pole prefers a smaller radius and longer length compared to other benchmark channels because the final state particles at the $Z$-pole have a more forward distribution. We also analyze the scaling behavior of the optimal tracker configuration at floating construction cost and observe a weak dependence.

physics.ins-det↗

arXiv2022

Boundary rigidity of Gromov hyperbolic spaces

We introduce the concept of boundary rigidity for Gromov hyperbolic spaces. We show that a proper geodesic Gromov hyperbolic space with a pole is boundary rigid if and only if its Gromov boundary is uniformly perfect. As an application, we show that for a non-compact Gromov hyperbolic complete Riemannian manifold or a Gromov hyperbolic uniform graph, boundary rigidity is equivalent to having positive Cheeger isoperimetric constant and also to being nonamenable. Moreover, several hyperbolic fillings of compact metric spaces are proved to be boundary rigid if and only if the metric spaces are uniformly perfect. Also, boundary rigidity is shown to be equivalent to being geodesically rich, a concept introduced by Shchur (J. Funct. Anal., 2013).

math.GT↗

arXiv2022

A Scalable Blockchain-based Smart Contract Model for Decentralized Voltage Stability Using Sharding Technique

Blockchain technologies are one possible avenue for increasing the resilience of the Smart Grid, by decentralizing the monitoring and control of system-level objectives such as voltage stability protection. They furthermore offer benefits in data immutability and traceability, as blockchains are cryptographically secured. However, the performance of blockchain-based systems in real-time grid monitoring and control has never been empirically tested. This study proposes implementing a decentralized voltage stability algorithm using blockchain-based smart contracts, as a testbed for evaluating the performance of blockchains in real-time control. We furthermore investigate sharding mechanisms as a means of improving the system's scalability with fixed computing resources. We implement our models as a proof-of-concept prototype system using Hyperledger Fabric as our blockchain platform, the Matpower library in MATLAB as our power system simulator, and Hyperledger Caliper as our performance evaluation tool. We found that sharding does indeed lead to a substantial improvement in system scalability for this domain, measured by both transaction success rates and transaction latency.

cs.DC↗

arXiv2022

Acoustic-Net: A Novel Neural Network for Sound Localization and Quantification

Acoustic source localization has been applied in different fields, such as aeronautics and ocean science, generally using multiple microphones array data to reconstruct the source location. However, the model-based beamforming methods fail to achieve the high-resolution of conventional beamforming maps. Deep neural networks are also appropriate to locate the sound source, but in general, these methods with complex network structures are hard to be recognized by hardware. In this paper, a novel neural network, termed the Acoustic-Net, is proposed to locate and quantify the sound source simply using the original signals. The experiments demonstrate that the proposed method significantly improves the accuracy of sound source prediction and the computing speed, which may generalize well to real data. The code and trained models are available at https://github.com/JoaquinChou/Acoustic-Net.

cs.SD↗

arXiv2022

Laser-induced electron Fresnel diffraction in the tunneling and over-barrier ionization

The photoelectron momentum distribution in the strong-field ionization has a variety of structures that reveal the complicated dynamics of this process. Recently, we identified a low-energy interference structure in the case of a super-intense extreme ultraviolet (XUV) laser pulse and attributed it to the laser-induced electron Fresnel diffraction. This structure is determined by the laser-induced electron displacement [Geng L et al. 2021 Phys. Rev. A 104(2) L021102]. In the present work, we find that the Fresnel diffraction picture is also present in the tunneling and over-barrier regime of ionization by short pulses. However, the electron displacement is now induced by the electric field component of the laser pulse, instead of by the magnetic field component in the case of the superintense XUV pulse. After corresponding modifications to our quantum and semiclassical models, we find the same physical mechanism of the Fresnel diffraction governs the low-energy interference structures along the laser polarization. The results predicted by the two models agree well with the accurate results from the numerical solution to the time-dependent Schrodinger equation.

physics.atom-ph↗

arXiv2022

Laser-induced electron Fresnel diffraction by XUV pulses at extreme intensity

Ionization of atoms and molecules in laser fields can lead to various interesting interference structures in the photoelectron spectrum. For the case of a super-intense extreme ultraviolet laser pulse, we identify a novel petal-like interference structure in the electron momentum distribution along the direction of the laser field propagation. We show that this structure is quite general and can be attributed to the Fresnel diffraction of the electronic wavepacket by the nucleus. Our results are demonstrated by numerically solving the time-dependent Schrodinger equation of the atomic hydrogen beyond the dipole approximation. By building an analytical model, we find that the electron displacement determines the aforementioned interference pattern. In addition, we establish the physical picture of laser-induced electron Fresnel diffraction which is reinforced by both quantum and semiclassical models.

physics.atom-ph↗

arXiv2022

RareGAN: Generating Samples for Rare Classes

We study the problem of learning generative adversarial networks (GANs) for a rare class of an unlabeled dataset subject to a labeling budget. This problem is motivated from practical applications in domains including security (e.g., synthesizing packets for DNS amplification attacks), systems and networking (e.g., synthesizing workloads that trigger high resource usage), and machine learning (e.g., generating images from a rare class). Existing approaches are unsuitable, either requiring fully-labeled datasets or sacrificing the fidelity of the rare class for that of the common classes. We propose RareGAN, a novel synthesis of three key ideas: (1) extending conditional GANs to use labelled and unlabelled data for better generalization; (2) an active learning approach that requests the most useful labels; and (3) a weighted loss function to favor learning the rare class. We show that RareGAN achieves a better fidelity-diversity tradeoff on the rare class than prior work across different applications, budgets, rare class fractions, GAN losses, and architectures.

cs.LG↗

arXiv2022

False Data Injection Attack on Electric Vehicle-Assisted Voltage Regulation

With the large scale penetration of electric vehicles (EVs) and the advent of bidirectional chargers, EV aggregators will become a major player in the voltage regulation market. This paper proposes a novel false data injection attack (FDIA) against the voltage regulation capacity estimation of EV charging stations, the process that underpins voltage regulation in distribution system. The proposed FDIA takes into account the uncertainty in EV mobility and network conditions. The attack vector with the largest expected adverse impact is the solution of a stochastic optimization problem subject to a constraint that ensures it can bypass bad data detection. We show that this attack vector can be determined by solving a sequence of convex quadratically constrained linear programs. The case studies examined in a co-simulation platform, based on two standard test feeders, reveal the vulnerability of the voltage regulation capacity estimation.

eess.SY↗

arXiv2021

A stochastic maximum principle for partially observed general mean-field control problems with only weak solution

In this paper we focus on a general type of mean-field stochastic control problem with partial observation, in which the coefficients depend in a non-linear way not only on the state process $X_t$ and its control $u_t$ but also on the conditional law $E[X_t|\mathcal{F}_t^Y]$ of the state process conditioned with respect to the past of observation process $Y$. We first deduce the well-posedness of the controlled system by showing weak existence and uniqueness in law. Neither supposing convexity of the control state space nor differentiability of the coefficients with respect to the control variable, we study Peng's stochastic maximum principle for our control problem. The novelty and the difficulty of our work stem from the fact that, given an admissible control $u$, the solution of the associated control problem is only a weak one. This has as consequence that also the probability measure in the solution $P^{u}=L^{u}_TQ$ depends on $u$ and has a density $L^{u}_T$ with respect to a reference measure $Q$. So characterizing an optimal control leads to the differentiation of non-linear functions $f(P^{u}\circ\{E^{P^{u}}[X_t|\mathcal{F}_t^Y]\}^{-1})$ with respect to $(L^{u}_T,X_t)$. This has as consequence for the study of Peng's maximum principle that we get a new type of first and second order variational equations and adjoint backward stochastic differential equations, all with new mean-field terms and with coefficients which are not Lipschitz. For their estimates and for those for the Taylor expansion new techniques have had to be introduced and rather technical results have had to be established. The necessary optimality condition we get extends Peng's one with new, non-trivial terms.

math.OC↗