arXiv subjects

Hu Li

Publications and source records attributed to Hu Li.

At least 19 recordsLinked to original sources

arXiv2025

Simplifying Root Cause Analysis in Kubernetes with StateGraph and LLM

Kubernetes, a notably complex and distributed system, utilizes an array of controllers to uphold cluster management logic through state reconciliation. Nevertheless, maintaining state consistency presents significant challenges due to unexpected failures, network disruptions, and asynchronous issues, especially within dynamic cloud environments. These challenges result in operational disruptions and economic losses, underscoring the necessity for robust root cause analysis (RCA) to enhance Kubernetes reliability. The development of large language models (LLMs) presents a promising direction for RCA. However, existing methodologies encounter several obstacles, including the diverse and evolving nature of Kubernetes incidents, the intricate context of incidents, and the polymorphic nature of these incidents. In this paper, we introduce SynergyRCA, an innovative tool that leverages LLMs with retrieval augmentation from graph databases and enhancement with expert prompts. SynergyRCA constructs a StateGraph to capture spatial and temporal relationships and utilizes a MetaGraph to outline entity connections. Upon the occurrence of an incident, an LLM predicts the most pertinent resource, and SynergyRCA queries the MetaGraph and StateGraph to deliver context-specific insights for RCA. We evaluate SynergyRCA using datasets from two production Kubernetes clusters, highlighting its capacity to identify numerous root causes, including novel ones, with high efficiency and precision. SynergyRCA demonstrates the ability to identify root causes in an average time of about two minutes and achieves an impressive precision of approximately 0.90.

cs.DC↗

arXiv2024

Dance of the ADS: Orchestrating Failures through Historically-Informed Scenario Fuzzing

As autonomous driving systems (ADS) advance towards higher levels of autonomy, orchestrating their safety verification becomes increasingly intricate. This paper unveils ScenarioFuzz, a pioneering scenario-based fuzz testing methodology. Designed like a choreographer who understands the past performances, it uncovers vulnerabilities in ADS without the crutch of predefined scenarios. Leveraging map road networks, such as OPENDRIVE, we extract essential data to form a foundational scenario seed corpus. This corpus, enriched with pertinent information, provides the necessary boundaries for fuzz testing in the absence of starting scenarios. Our approach integrates specialized mutators and mutation techniques, combined with a graph neural network model, to predict and filter out high-risk scenario seeds, optimizing the fuzzing process using historical test data. Compared to other methods, our approach reduces the time cost by an average of 60.3%, while the number of error scenarios discovered per unit of time increases by 103%. Furthermore, we propose a self-supervised collision trajectory clustering method, which aids in identifying and summarizing 54 high-risk scenario categories prone to inducing ADS faults. Our experiments have successfully uncovered 58 bugs across six tested systems, emphasizing the critical safety concerns of ADS.

cs.AI↗

arXiv2023

A fast compact difference scheme with unequal time-steps for the tempered time-fractional Black-Scholes model

The Black-Scholes (B-S) equation has been recently extended as a kind of tempered time-fractional B-S equations, which becomes an interesting mathematical model in option pricing. In this study, we provide a fast numerical method to approximate the solution of the tempered time-fractional B-S model. To achieve high-order accuracy in space and overcome the weak initial singularity of exact solution, we combine the compact difference operator with L1-type approximation under nonuniform time steps to yield the numerical scheme. The convergence of the proposed difference scheme is proved to be unconditionally stable. Moreover, the kernel function in the tempered Caputo fractional derivative is approximated by sum-of-exponentials, which leads to a fast unconditionally stable compact difference method that reduces the computational cost. Finally, numerical results demonstrate the effectiveness of the proposed methods.

math.NA↗

arXiv2021

On the bilateral preconditioning for an L2-type all-at-once system arising from time-space fractional Bloch-Torrey equations

Time-space fractional Bloch-Torrey equations (TSFBTEs) are developed by some researchers to investigate the relationship between diffusion and fractional-order dynamics. In this paper, we first propose a second-order implicit difference scheme for TSFBTEs by employing the recently proposed L2-type formula [A.~A.~Alikhanov, C.~Huang, Appl.~Math.~Comput.~(2021) 126545]. Then, we prove the stability and the convergence of the proposed scheme. Based on such a numerical scheme, an L2-type all-at-once system is derived. In order to solve this system in a parallel-in-time pattern, a bilateral preconditioning technique is designed to accelerate the convergence of Krylov subspace solvers according to the special structure of the coefficient matrix of the system. We theoretically show that the condition number of the preconditioned matrix is uniformly bounded by a constant for the time fractional order $\alpha \in (0,0.3624)$. Numerical results are reported to show the efficiency of our method.

math.NA↗

arXiv2019

Predicting Heart Failure Readmission from Clinical Notes Using Deep Learning

Heart failure hospitalization is a severe burden on healthcare. How to predict and therefore prevent readmission has been a significant challenge in outcomes research. To address this, we propose a deep learning approach to predict readmission from clinical notes. Unlike conventional methods that use structured data for prediction, we leverage the unstructured clinical notes to train deep learning models based on convolutional neural networks (CNN). We then use the trained models to classify and predict potentially high-risk admissions/patients. For evaluation, we trained CNNs using the discharge summary notes in the MIMIC III database. We also trained regular machine learning models based on random forest using the same datasets. The result shows that deep learning models outperform the regular models in prediction tasks. CNN method achieves a F1 score of 0.756 in general readmission prediction and 0.733 in 30-day readmission prediction, while random forest only achieves a F1 score of 0.674 and 0.656 respectively. We also propose a chi-square test based method to interpret key features associated with deep learning predicted readmissions. It reveals clinical insights about readmission embedded in the clinical notes. Collectively, our method can make the human evaluation process more efficient and potentially facilitate the reduction of readmission rates.

cs.CL↗

arXiv2016

Debugging OpenStack Problems Using a State Graph Approach

It is hard to operate and debug systems like OpenStack that integrate many independently developed modules with multiple levels of abstractions. A major challenge is to navigate through the complex dependencies and relationships of the states in different modules or subsystems, to ensure the correctness and consistency of these states. We present a system that captures the runtime states and events from the entire OpenStack-Ceph stack, and automatically organizes these data into a graph that we call system operation state graph (SOSG).With SOSG we can use intuitive graph traversal techniques to solve problems like reasoning about the state of a virtual machine. Also, using graph-based anomaly detection, we can automatically discover hidden problems in OpenStack. We have a scalable implementation of SOSG, and evaluate the approach on a 125-node production OpenStack cluster, finding a number of interesting problems.

cs.DC↗

arXiv2014

Binocular parallax stereo imaging based on correlation matching algorithm

The intensity fluctuation correlation of pseudo-thermal light can be utilized to realize binocular parallax stereo imaging (BPSI). With the help of correlation matching algorithm, the matching precision of feature points can reach one pixel authentically. The implementations of the proposed BPSI system with real objects were demonstrated in detail. And the experimental results indicated that the proposed system performs better when the object's superficial characteristics are not obvious, for example its surface reflectivity is constant.

physics.optics↗

arXiv2014

Epigenetic landscapes explain partially reprogrammed cells and identify key reprogramming genes

A common metaphor for describing development is a rugged "epigenetic landscape" where cell fates are represented as attracting valleys resulting from a complex regulatory network. Here, we introduce a framework for explicitly constructing epigenetic landscapes that combines genomic data with techniques from spin-glass physics. Each cell fate is a dynamic attractor, yet cells can change fate in response to external signals. Our model suggests that partially reprogrammed cells are a natural consequence of high-dimensional landscapes, and predicts that partially reprogrammed cells should be hybrids that co-express genes from multiple cell fates. We verify this prediction by reanalyzing existing datasets. Our model reproduces known reprogramming protocols and identifies candidate transcription factors for reprogramming to novel cell fates, suggesting epigenetic landscapes are a powerful paradigm for understanding cellular identity.

q-bio.MN↗

arXiv2014

Analysis of Capacity Region of Delay-Tolerant Hybrid Mobile Ad Hoc Networks

Network capacity region is an important character of mobile ad hoc networks. Using cell-partitioned model, an expression of upper bound of delay-tolerant hybrid mobile ad hoc network is deduced regardless of coverage of base stations, types of mobile process, scheduling and routing algorithms. The limitation of the upper bound is derived, Analysis of the limitation of upper bound is carried out when the steady-state follows even-distribution law. The relationship among limitation of capacity, the node density and coverage of base station is analyzed.

cs.NI↗

arXiv2013

A Wrapper of PCI Express with FIFO Interfaces based on FPGA

This paper proposes a PCI Express (PCIE) Wrapper core named PWrapper with FIFO interfaces. Compared with other PCIE solutions, PWrapper has several advantages such as flexibility, isolation of clock domain, etc. PWrapper is implemented and verified on Vertex -5-FX70T which is a development board provided by Xilinx Inc. Architecture of PWrapper and design of two key modules are illustrated, which timing optimization methods have been adopted. Then we explained the advantages and challenges of on-chip interfaces technology based on FIFOs. The verification results show that PWrapper can achieve the speed of 1.8Gbps (Giga bits per second).

cs.AR↗

arXiv2005

Relativistic Model of Triquark Structure

At this point it is still unclear whether pentaquarks exist. While they have be seen in some experiments there are many experiments in which they are not found. On the assumption that pentaquarks exist, several authors have studied the properties of pentaquarks. One description considered is that of pentaquarks which consist of a diquark coupled to a triquark. There is a quite extensive literature concerning the properties of diquarks and their importance in the description of the nucleon has been considered by several authors. On the other hand, there is little work reported concerning the description of triquarks. In the present work we study a model for the triquark in which it is composed of a component which contains a quark coupled to a scalar diquark and another two components in which there is a quark coupled to a kaon. We solve for the wave function of the triquark and obtain a mass for the triquark of 0.81 GeV which is quite close to the value of 0.80 GeV obtained in a QCD sum rule study of triquark properties.

nucl-th↗

arXiv2005

Relativistic Calculation of Pentaquark Widths

We calculate the widths of the various pentaquarks in a relativistic model in which the pentaquark is considered to be composed of a scalar diquark and a spin 1/2 triquark. We consider both positive and negative parity for the pentaquark. There is a single parameter in our model which we vary and which describes the size of the pentaquark. We obtain quite small widths for the decay Theta^(+) -> N+K^(+) and for Theta_c^0 -> P+D^{*-} consistent with the experimental situation. For the sum of the decay widths for Xi(bar)^(--) -> Xi^(-) + pi^(+) and Xi(bar)^(--) -> Sigma^(-) + K^(-) we find values of the order of 4-8 MeV for pentaquarks of the characteristic size considered in this work. (The experimental situation with respect to te observation of the Xi(bar)^(--) is somewhat uncertain at this time.) We also provide results for the decays N^(+) -> N + pi and N_s^(+) -> Lambda^(0) + K^(+). Our model of confinement plays an important role in our analysis and makes it possible to use Feynman diagrams to describe the decay of the pentaquark.

hep-ph↗

arXiv2005

Relativistic Calculation of the Width of the Theta (1540) Pentaquark

We calculate the width of the Theta(1540) pentaquark in a relativistic model in which the pentaquark is considered to be composed of a scalar diquark and a spin 1/2 triquark. We consider both positive and negative parity for the pentaquark. There is a single parameter in our model which we vary and which describes the size of the pentaquark. If the pentaquark size is somewhat smaller than that of the nucleon, we find quite small widths for the pentaquark of about 1 MeV or less. Our model of confinement plays an important role in our analysis and makes it possible to use Feynman diagrams to describe the decay of the pentaquark.

hep-ph↗

arXiv2004

Calculation of the Momentum Dependence of Hadronic Current Correlation Functions at Finite Temperature

We have calculated spectral functions associated with hadronic current correlation functions for vector currents at finite temperature. We made use of a model with chiral symmetry, temperature-dependent coupling constants and temperature-dependent momentum cutoff parameters. Our model has two parameters which are used to fix the magnitude and position of the large peak seen in the spectral functions. In our earlier work, good fits were obtained for the spectral functions that were extracted from lattice data by means of the maximum entropy method (MEM). In the present work we extend our calculations and provide values for the three-momentum dependence of the vector correlation function at T=1.5T_c. These results are used to obtain the correlation function in coordinate space, which is usually parametrized in terms of a screening mass. Our results for the three-momentum dependence of the spectral functions are similar to those found in a recent lattice QCD calculation for charmonium [S. Datta, F. Karsch, P. Petreczky and I. Wetzorke, hep-lat/0312037]. For a limited range we find the exponential behavior in coordinate space that is usually obtained for the spectral function for T>T_c and which allows for the definition of a screening mass.

nucl-th↗

arXiv2004

Calculation of Screening Masses in a Chiral Quark Model

We consider a simple model for the coordinate-space vacuum polarization function which is often parametrized in terms of a screening mass. We discuss the circumstances in which the standard result for the screening mass, $m_{sc}=πT$, is obtained. In the model considered here, that result is obtained when the momenta in the relevant vacuum polarization integral are small with respect to the first Matsubara frequency.

nucl-th↗

arXiv2004

Quark Propagation in the Quark-Gluon Plasma

It has recently been suggested that the quark-gluon plasma formed in heavy-ion collisions behaves as a nearly ideal fluid. That behavior may be understood if the quark and antiquark mean-free- paths are very small in the system, leading to a "sticky molasses" description of the plasma, as advocated by the Stony Brook group. This behavior may be traced to the fact that there are relatively low-energy $q\bar{q}$ resonance states in the plasma leading to very large scattering lengths for the quarks. These resonances have been found in lattice simulation of QCD using the maximum entropy method (MEM). We have used a chiral quark model, which provides a simple representation of effects due to instanton dynamics, to study the resonances obtained using the MEM scheme. In the present work we use our model to study the optical potential of a quark in the quark-gluon plasma and calculate the quark mean-free-path. Our results represent a specific example of the dynamics of the plasma as described by the Stony Brook group.

hep-ph↗

arXiv2004

Chiral Quark Model Calculation of the Momentum Dependence of Hadronic Current Correlation Functions at Finite Temperature

We calculate spectral functions associated with hadronic current correlation functions for vector currents at finite temperature. We make use of a model with chiral symmetry, temperature-dependent coupling constants and temperature-dependent momentum cutoff parameters. Our model has two parameters which are used to fix the magnitude and position of the large peak seen in the spectral functions. In our earlier work, good fits were obtained for the spectral functions that were extracted from lattice data by means of the maximum entropy method (MEM). In the present work we extend our calculations and provide values for the three-momentum dependence of the vector correlation function at $T=1.5T_c$. These results are used to obtain the correlation function in coordinate space, which is usually parametrized in terms of a screening mass. Our results for the three-momentum dependence of the spectral functions are similar to those found in a recent lattice QCD calculation for charmonium [S. Datta, F. Karsch, P. Petreczky and I. Wetzorke, hep-lat/0312037]. However, we do not find the expontential behavior in coordinate space that is usually assumed for the spectral function for $T>T_c$ and which allows for the definition of a screening mass.

hep-ph↗

arXiv2003

Quark Model Calculations of Spectral Functions of Hadronic Current Correlation Functions at Finite Temperature

We calculate spectral functions associated with hadronic current correlation functions for vector and pseudoscalar currents at finite temperature. We make use of the Nambu--Jona--Lasinio (NJL) model with temperature-dependent coupling constants and temperature-dependent momentum cutoff parameters. At low energies, good fits are obtained for the spectral functions that were extracted from lattice data by means of the maximum entropy method (MEM). Our model has two parameters which are used to fix the magnitude and position of the large peak seen in the spectral functions. With those two parameters fixed, we obtain a satisfactory fit to the width of the peak. The model then also reproduces the energy of a second peak seen in the spectral functions. In the case of the pseudoscalar spectral function, the calculated peak is about 20 percent higher than that found for the spectral function obtained from the lattice data. However, it appears that the second peak is a lattice artifact [ P. Petreczky, private communication ] and our fit to the second peak may not be meaningful. We conclude that the NJL model may have a broader range of application than previously considered to be the case, if one allows for significant temperature dependence of the parameters of the model, as well as rather large values of the momentum cutoff parameter. Our treatment of temperature-dependent coupling constants and cutoff parameters is analogous to the procedure introduced by R. Casalbuoni, R. Gatto, G. Nardulli, and M. Ruggieri, [ Phys. Rev. D \textbf{68}, 034024 (2003) ], who make use of the NJL model at finite density and find that they need to use the density-dependent coupling constants and density -dependent cutoff parameters to study matter at high density.

hep-ph↗