arXiv subjects

Shuo Yuan

Publications and source records attributed to Shuo Yuan.

At least 37 records · Page 2Linked to original sources

arXiv2024

Stochastic Hybrid System Modeling and State Estimation of Modern Power Systems under Contingency

This paper introduces a stochastic hybrid system (SHS) framework in state space model to capture sensor, communication, and system contingencies in modern power systems (MPS). Within this new framework, the paper concentrates on the development of state estimation methods and algorithms to provide reliable state estimation under randomly intermittent and noisy sensor data. MPSs employ diversified measurement devices for monitoring system operations that are subject to random measurement errors and rely on communication networks to transmit data whose channels encounter random packet loss and interruptions. The contingency and noise form two distinct and interacting stochastic processes that have a significant impact on state estimation accuracy and reliability. This paper formulates stochastic hybrid system models for MPSs, introduces coordinated observer design algorithms for state estimation, and establishes their convergence and reliability properties. A further study reveals a fundamental design tradeoff between convergence rates and steady-state error variances. Simulation studies on the IEEE 5-bus system and IEEE 33-bus system are used to illustrate the modeling methods, observer design algorithms, convergence properties, performance evaluations, and impact sensor system selections.

eess.SY↗

arXiv2024

Joint Beam Direction Control and Radio Resource Allocation in Dynamic Multi-beam LEO Satellite Networks

Multi-beam low earth orbit (LEO) satellites are emerging as key components in beyond 5G and 6G to provide global coverage and high data rate. To fully unleash the potential of LEO satellite communication, resource management plays a key role. However, the uneven distribution of users, the coupling of multi-dimensional resources, complex inter-beam interference, and time-varying network topologies all impose significant challenges on effective communication resource management. In this paper, we study the joint optimization of beam direction and the allocation of spectrum, time, and power resource in a dynamic multi-beam LEO satellite network. The objective is to improve long-term user sum data rate while taking user fairness into account. Since the concerned resource management problem is mixed-integer non-convex programming, the problem is decomposed into three subproblems, namely beam direction control and time slot allocation, user subchannel assignment, and beam power allocation. Then, these subproblems are solved iteratively by leveraging matching with externalities and successive convex approximation, and the proposed algorithms are analyzed in terms of stability, convergence, and complexity. Extensive simulations are conducted, and the results demonstrate that our proposal can improve the number of served users by up to two times and the sum user data rate by up to 68%, compared to baseline schemes.

cs.IT↗

arXiv2024

A New Creative Generation Pipeline for Click-Through Rate with Stable Diffusion Model

In online advertising scenario, sellers often create multiple creatives to provide comprehensive demonstrations, making it essential to present the most appealing design to maximize the Click-Through Rate (CTR). However, sellers generally struggle to consider users preferences for creative design, leading to the relatively lower aesthetics and quantities compared to Artificial Intelligence (AI)-based approaches. Traditional AI-based approaches still face the same problem of not considering user information while having limited aesthetic knowledge from designers. In fact that fusing the user information, the generated creatives can be more attractive because different users may have different preferences. To optimize the results, the generated creatives in traditional methods are then ranked by another module named creative ranking model. The ranking model can predict the CTR score for each creative considering user features. However, the two above stages are regarded as two different tasks and are optimized separately. In this paper, we proposed a new automated Creative Generation pipeline for Click-Through Rate (CG4CTR) with the goal of improving CTR during the creative generation stage. Our contributions have 4 parts: 1) The inpainting mode in stable diffusion is firstly applied to creative generation task in online advertising scene. A self-cyclic generation pipeline is proposed to ensure the convergence of training. 2) Prompt model is designed to generate individualized creatives for different user groups, which can further improve the diversity and quality. 3) Reward model comprehensively considers the multimodal features of image and text to improve the effectiveness of creative ranking task, and it is also critical in self-cyclic pipeline. 4) The significant benefits obtained in online and offline experiments verify the significance of our proposed method.

cs.IR↗

arXiv2023

Joint Network Function Placement and Routing Optimization in Dynamic Software-defined Satellite-Terrestrial Integrated Networks

Software-defined satellite-terrestrial integrated networks (SDSTNs) are seen as a promising paradigm for achieving high resource flexibility and global communication coverage. However, low latency service provisioning is still challenging due to the fast variation of network topology and limited onboard resource at low earth orbit satellites. To address this issue, we study service provisioning in SDSTNs via joint optimization of virtual network function (VNF) placement and routing planning with network dynamics characterized by a time-evolving graph. Aiming at minimizing average service latency, the corresponding problem is formulated as an integer nonlinear programming under resource, VNF deployment, and time-slotted flow constraints. Since exhaustive search is intractable, we transform the primary problem into an integer linear programming by involving auxiliary variables and then propose a Benders decomposition based branch-and-cut (BDBC) algorithm. Towards practical use, a time expansion-based decoupled greedy (TEDG) algorithm is further designed with rigorous complexity analysis. Extensive experiments demonstrate the optimality of BDBC algorithm and the low complexity of TEDG algorithm. Meanwhile, it is indicated that they can improve the number of completed services within a configuration period by up to 58% and reduce the average service latency by up to 17% compared to baseline schemes.

eess.SP↗

arXiv2023

Exploring Model Dynamics for Accumulative Poisoning Discovery

Adversarial poisoning attacks pose huge threats to various machine learning applications. Especially, the recent accumulative poisoning attacks show that it is possible to achieve irreparable harm on models via a sequence of imperceptible attacks followed by a trigger batch. Due to the limited data-level discrepancy in real-time data streaming, current defensive methods are indiscriminate in handling the poison and clean samples. In this paper, we dive into the perspective of model dynamics and propose a novel information measure, namely, Memorization Discrepancy, to explore the defense via the model-level information. By implicitly transferring the changes in the data manipulation to that in the model outputs, Memorization Discrepancy can discover the imperceptible poison samples based on their distinct dynamics from the clean samples. We thoroughly explore its properties and propose Discrepancy-aware Sample Correction (DSC) to defend against accumulative poisoning attacks. Extensive experiments comprehensively characterized Memorization Discrepancy and verified its effectiveness. The code is publicly available at: https://github.com/tmlr-group/Memorization-Discrepancy.

cs.LG↗

arXiv2023

Tracking performance of PID for nonlinear stochastic systems

In this paper, we will consider a class of continuous-time stochastic control systems with both unknown nonlinear structure and unknown disturbances, and investigate the capability of the classical proportional-integral-derivative(PID) controller in tracking time-varying reference signals. First, under some suitable conditions on system nonlinear functions, reference signals, and unknown disturbances, we will show that PID controllers can be designed to globally stabilize such systems and ensure the boundedness of the tracking error. Analytic design formulae for PID gain matrices are also provided, which only involve some prior knowledge of the partial derivatives of system structural nonlinear functions. Besides, it will be shown that the steady-state tracking error hinges on three critical factors: i) the change rate of reference signals and external disturbances; ii) the intensity of random noises; iii) the selection of PID gains, and can be made arbitrarily small by choosing PID gains suitably large. Finally, by introducing a desired transient process which is shaped from the reference signal, we will present a new PID tuning rule, which can guarantee both nice steady-state and superior transient control performances.

eess.SY↗

arXiv2022

Cosmological Studies from HSC-SSP Tomographic Weak Lensing Peak Abundances

We perform weak lensing tomographic peak studies using the first-year shear data from Hyper Suprime-Cam Subaru Strategic Program (HSC-SSP) survey. The effective area used in our analyses after field selection, mask and boundary exclusions is $\sim 58 °^2$. The source galaxies are divided into low- and high-redshift bins with $0.2\le z_p\le0.85$ and $0.85\le z_p\le1.5$, respectively. We utilize our halo-based theoretical peak model including the projection effect of large-scale structures to derive cosmological constraints from the observed tomographic high peak abundances with the signal-to-noise ratio in the range of $ν_{\rm N}=[3.5,5.5]$. These high peaks are closely associated with the lensing effects of massive clusters of galaxies. Thus the inclusion of their member galaxies in the shear catalog can lead to significant source clustering and dilute their lensing signals. We account for this systematic effect in our theoretical modelling. Additionally, the impacts of baryonic effects, galaxy intrinsic alignments, as well as residual uncertainties in shear and photometric redshift calibrations are also analyzed. Within the flat $Λ$CDM model, the derived constraint is $S_8=0.758_{-0.076}^{+0.033}$ and $0.768_{-0.057}^{+0.030}$ with the source clustering information measured from the two cluster catalogs, CAMIRA and WZL, respectively. The asymmetric uncertainties are due to the different degeneracy direction of $(Ω_{\rm m}, σ_8)$ from high peak abundances comparing to that from the cosmic shear two-point correlations which give rise approximately the power index $α=0.5$. Fitting to our constraints, we obtain $α\approx 0.38$ and $Σ_8=0.772_{-0.032}^{+0.028}$ (CAMIRA) and $0.781_{-0.033}^{+0.028}$ (WZL). In comparison with the results from non-tomographic peak analyses, the $1σ$ uncertainties on $Σ_8$ are reduced by a factor of $\sim1.3$.

astro-ph.CO↗

arXiv2022

ROI-Constrained Bidding via Curriculum-Guided Bayesian Reinforcement Learning

Real-Time Bidding (RTB) is an important mechanism in modern online advertising systems. Advertisers employ bidding strategies in RTB to optimize their advertising effects subject to various financial requirements, especially the return-on-investment (ROI) constraint. ROIs change non-monotonically during the sequential bidding process, and often induce a see-saw effect between constraint satisfaction and objective optimization. While some existing approaches show promising results in static or mildly changing ad markets, they fail to generalize to highly dynamic ad markets with ROI constraints, due to their inability to adaptively balance constraints and objectives amidst non-stationarity and partial observability. In this work, we specialize in ROI-Constrained Bidding in non-stationary markets. Based on a Partially Observable Constrained Markov Decision Process, our method exploits an indicator-augmented reward function free of extra trade-off parameters and develops a Curriculum-Guided Bayesian Reinforcement Learning (CBRL) framework to adaptively control the constraint-objective trade-off in non-stationary ad markets. Extensive experiments on a large-scale industrial dataset with two problem settings reveal that CBRL generalizes well in both in-distribution and out-of-distribution data regimes, and enjoys superior learning efficiency and stability.

cs.LG↗

arXiv2022

A new consistency test for $Λ$CDM cosmology using galaxy surveys

We propose a new consistency test for the $Λ$CDM cosmology using baryonic acoustic oscillations (BAO) and redshift space distortion (RSD) measurements from galaxy redshift surveys. Specifically, we determine the peak position of $fσ_8(z)$ in redshift $z$ offered by a RSD measurement, and compare it to the one predicted by the BAO observables assuming a flat $Λ$CDM cosmology. We demonstrate this new test using the simulated data for the DESI galaxy survey, and argue that this test complements those using the background observables alone, and is less subject to systematics in the RSD analysis, compared to traditional methods using values of $fσ_8(z)$ directly.

astro-ph.CO↗

arXiv2021

Joint Sensing, Communication, and Computation Resource Allocation for Cooperative Perception in Fog-Based Vehicular Networks

To enlarge the perception range and reliability of individual autonomous vehicles, cooperative perception has been received much attention. However, considering the high volume of shared messages, limited bandwidth and computation resources in vehicular networks become bottlenecks. In this paper, we investigate how to balance the volume of shared messages and constrained resources in fog-based vehicular networks. To this end, we first characterize sum satisfaction of cooperative perception taking account of its spatial-temporal value and latency performance. Next, the sensing block message, communication resource block, and computation resource are jointly allocated to maximize the sum satisfaction of cooperative perception, while satisfying the maximum latency and sojourn time constraints of vehicles. Owing to its non-convexity, we decouple the original problem into two separate sub-problems and devise corresponding solutions. Simulation results demonstrate that our proposed scheme can effectively boost the sum satisfaction of cooperative perception compared with existing baselines.

cs.NI↗

arXiv2021

Exploration in Online Advertising Systems with Deep Uncertainty-Aware Learning

Modern online advertising systems inevitably rely on personalization methods, such as click-through rate (CTR) prediction. Recent progress in CTR prediction enjoys the rich representation capabilities of deep learning and achieves great success in large-scale industrial applications. However, these methods can suffer from lack of exploration. Another line of prior work addresses the exploration-exploitation trade-off problem with contextual bandit methods, which are recently less studied in the industry due to the difficulty in extending their flexibility with deep models. In this paper, we propose a novel Deep Uncertainty-Aware Learning (DUAL) method to learn CTR models based on Gaussian processes, which can provide predictive uncertainty estimations while maintaining the flexibility of deep neural networks. DUAL can be easily implemented on existing models and deployed in real-time systems with minimal extra computational overhead. By linking the predictive uncertainty estimation ability of DUAL to well-known bandit algorithms, we further present DUAL-based Ad-ranking strategies to boost up long-term utilities such as the social welfare in advertising systems. Experimental results on several public datasets demonstrate the effectiveness of our methods. Remarkably, an online A/B test deployed in the Alibaba display advertising platform shows an 8.2% social welfare improvement and an 8.0% revenue lift.

cs.IR↗

arXiv2020

Performance Analysis of Reversible Binding Receptor Based Decode-and-Forward Relay in Molecular Communication Systems

Molecular communication (MC) allows nanomachines to communicate and cooperate with each other in a fluid environment. The diffusion-based MC is popular but is easily constrained by the transmit distance due to the severe attenuation of molecule concentrations. In this letter, we present a decode-and-forward (DF) relay strategy for the reversible binding receptor in the diffusion-based MC system. The time-varying spatial distribution of the information molecules based on the reversible association and dissociation between ligand and receptor at the surface of receiver is characterized. An analytical expression for the evaluation of expected error probability is derived, and the key factors impacting on the performance are exploited. Results show that with a constant molecular budget, the proposal can improve the performance significantly, and the performance gain can be enhanced by optimizing the position of the relay node and the number of molecules assigned to the source node.

cs.IT↗

arXiv2019

Cosmological studies from tomographic weak lensing peak abundances and impacts of photo-z errors

Weak lensing peak abundance analyses have been applied in different surveys and demonstrated to be a powerful statistics in extracting cosmological information complementary to cosmic shear two-point correlation studies. Future large surveys with high number densities of galaxies enable tomographic peak analyses. Focusing on high peaks, we investigate quantitatively how the tomographic redshift binning can enhance the cosmological gains. We also perform detailed studies about the degradation of cosmological information due to photometric redshift (photo-z) errors. We show that for surveys with the number density of galaxies $\sim40\,{\rm arcmin^{-2}}$, the median redshift $\sim1$, and the survey area of $\sim15000\,{\rm deg^{2}}$, the 4-bin tomographic peak analyses can reduce the error contours of $(Ω_{\rm m},σ_{8})$ by a factor of $5$ comparing to 2-D peak analyses in the ideal case of photo-z error being absent. More redshift bins can hardly lead to significantly better constraints. The photo-z error model here is parametrized by $z_{\rm bias}$ and $σ_{\rm ph}$ and the fiducial values of $z_{\rm bias}=0.003$ and $σ_{\rm ph}=0.02$ is taken. We find that using tomographic peak analyses can constrain the photo-z errors simultaneously with cosmological parameters. For 4-bin analyses, we can obtain $σ(z_{\rm bias})/z_{\rm bias}\sim10\%$ and $σ(σ_{\rm ph})/σ_{\rm ph}\sim5\%$ without assuming priors on them. Accordingly, the cosmological constraints on $Ω_{\rm m}$ and $σ_{8}$ degrade by a factor of $\sim2.2$ and $\sim1.8$, respectively, with respect to zero uncertainties on photo-z parameters. We find that the uncertainty of $z_{\rm bias}$ plays more significant roles in degrading the cosmological constraints than that of $σ_{\rm ph}$.

astro-ph.CO↗

arXiv2018

Deep CFHT Y-band imaging of VVDS-F22 field: I. data products and photometric redshifts

We present our deep $Y$-band imaging data of a two square degree field within the F22 region of the VIMOS VLT Deep Survey. The observations were conducted using the WIRCam instrument mounted at the Canada--France--Hawaii Telescope (CFHT). The total on-sky time was 9 hours, distributed uniformly over 18 tiles. The scientific goals of the project are to select faint quasar candidates at redshift $z>2.2$, and constrain the photometric redshifts for quasars and galaxies. In this paper, we present the observation and the image reduction, as well as the photometric redshifts that we derived by combining our $Y$-band data with the CFHTLenS $u^*g'r'i'z'$ optical data and UKIDSS DXS $JHK$ near-infrared data. With $J$-band image as reference total $\sim$80,000 galaxies are detected in the final mosaic down to $Y$-band $5σ$ point source limiting depth of 22.86 mag. Compared with the $\sim$3500 spectroscopic redshifts, our photometric redshifts for galaxies with $z<1.5$ and $i'\lesssim24.0$ mag have a small systematic offset of $|Δ{z}|\lesssim0.2$, 1$σ$ scatter $0.03<σ_{Δz} < 0.06$, and less than 4.0% of catastrophic failures. We also compare to the CFHTLenS photometric redshifts, and find that ours are more reliable at $z\gtrsim0.6$ because of the inclusion of the near-infrared bands. In particular, including the $Y$-band data can improve the accuracy at $z\sim 1.0-2.0$ because the location of the 4000Å-break is better constrained. The $Y$-band images, the multi-band photometry catalog and the photometric redshifts are released at \url{http://astro.pku.edu.cn/astro/data/DYI.html}.

astro-ph.GA↗

arXiv2018

The Correspondence between Convergence Peaks from Weak Lensing and Massive Dark Matter Haloes

The convergence peaks, constructed from galaxy shape measurement in weak lensing, is a powerful probe of cosmology as the peaks can be connected with the underlined dark matter haloes. However the capability of convergence peak statistic is affected by the noise in galaxy shape measurement, signal to noise ratio as well as the contribution from the projected mass distribution from the large-scale structures along the line of sight (LOS). In this paper we use the ray-tracing simulation on a curved sky to investigate the correspondence between the convergence peak and the dark matter haloes at the LOS. We find that, in case of no noise and for source galaxies at $z_{\rm s}=1$, more than $65\%$ peaks with $\text{SNR} \geq 3$ (signal to noise ratio) are related to more than one massive haloes with mass larger than $10^{13} {\rm M}_{\odot}$. Those massive haloes contribute $87.2\%$ to high peaks ($\text{SNR} \geq 5$) with the remaining contributions are from the large-scale structures. On the other hand, the peaks distribution is skewed by the noise in galaxy shape measurement, especially for lower SNR peaks. In the noisy field where the shape noise is modelled as a Gaussian distribution, about $60\%$ high peaks ($\text{SNR} \geq 5$) are true peaks and the fraction decreases to $20\%$ for lower peaks ($ 3 \leq \text{SNR} < 5$). Furthermore, we find that high peaks ($\text{SNR} \geq 5$) are dominated by very massive haloes larger than $10^{14} {\rm M}_{\odot}$.

astro-ph.CO↗

arXiv2018

Projection effects of large-scale structures on weak-lensing peak abundances

High peaks in weak lensing (WL) maps originate dominantly from the lensing effects of single massive halos. Their abundance is therefore closely related to the halo mass function and thus a powerful cosmological probe. On the other hand, however, besides individual massive halos, large-scale structures (LSS) along lines of sight also contribute to the peak signals. In this paper, with ray tracing simulations, we investigate the LSS projection effects. We show that for current surveys with a large shape noise, the stochastic LSS effects are subdominant. For future WL surveys with source galaxies having a median redshift $z_{\mathrm{med}}\sim1$ or higher, however, they are significant. For the cosmological constraints derived from observed WL high peak counts, severe biases can occur if the LSS effects are not taken into account properly. We extend the model of \citet{Fan2010} by incorporating the LSS projection effects into the theoretical considerations. By comparing with simulation results, we demonstrate the good performance of the improved model and its applicability in cosmological studies.

astro-ph.CO↗

arXiv2018

Deep CFHT Y-band Imaging of VVDS-F22 Field: II. Quasar Selection and Quasar Luminosity Function

We report the result of a faint quasar survey in a one square degree field. The aim is to test the Y-K/g-z and J-K/i-Y color selection criteria for quasars at faint magnitude, to obtain a complete sample of quasars based on deep optical and near-infrared color-color selection, and to measure the faint end of quasar luminosity function (QLF) over a wide redshift range. We carried out a quasar survey based on the Y-K/g-z and J-K/i-Y quasar selection criteria, using the deep Y-band data obtained from our CFHT/WIRCam Y-band images in a two-degree field within the F22 field of the VIMOS VLT deep survey, optical co-added data from Sloan Digital Sky Survey Stripe 82 and deep near-infrared data from the UKIDSS Deep Extragalactic Survey in the same field. We discovered 25 new quasars at 0.5 < z < 4.5 and i < 22.5 mag within one square degree field. The survey significantly increases the number of faint quasars in this field, especially at z ~ 2-3. It confirms that our color selections are highly complete in a wide redshift range (z < 4.5), especially over the quasar number density peak at z ~ 2-3, even for faint quasars. Combining all previous known quasars and new discoveries, we construct a sample with 109 quasars, and measure the binned QLF and parametric QLF. Although the sample is small, our results agree with a pure luminosity evolution at lower redshift and luminosity evolution and density evolution model at redshift z > 2.5.

astro-ph.GA↗

arXiv2017

Cosmological neutrino simulations at extreme scale

Constraining neutrino mass remains an elusive challenge in modern physics. Precision measurements are expected from several upcoming cosmological probes of large-scale structure. Achieving this goal relies on an equal level of precision from theoretical predictions of neutrino clustering. Numerical simulations of the non-linear evolution of cold dark matter and neutrinos play a pivotal role in this process. We incorporate neutrinos into the cosmological N-body code CUBEP3M and discuss the challenges associated with pushing to the extreme scales demanded by the neutrino problem. We highlight code optimizations made to exploit modern high performance computing architectures and present a novel method of data compression that reduces the phase-space particle footprint from 24 bytes in single precision to roughly 9 bytes. We scale the neutrino problem to the Tianhe-2 supercomputer and provide details of our production run, named TianNu, which uses 86% of the machine (13,824 compute nodes). With a total of 2.97 trillion particles, TianNu is currently the world's largest cosmological N-body simulation and improves upon previous neutrino simulations by two orders of magnitude in scale. We finish with a discussion of the unanticipated computational challenges that were encountered during the TianNu runtime.

astro-ph.CO↗