检索结果-南通市图书馆

Enhancing cut selection through reinforcement learning

在线全文

学校读者我要写书评

暂无评论

Science China Mathematics 2024年第6期67卷 1377-1394页

作者： Shengchao Wang Liang Chen Lingfeng Niu Yu-Hong Dai academy of Mathematics and Systems Science Chinese Academy of SciencesBeijing100190China Research Center on Fictitious Economy and Data Science Chinese Academy of SciencesBeijing100190China

With the rapid development of artificial intelligence in recent years,applying various learning techniques to solve mixed-integer linear programming(MILP)problems has emerged as a burgeoning research domain.Apart from constructing end-to-end models directly,integrating learning approaches with some modules in the traditional methods for solving MILPs is also a promising direction.The cutting plane method is one of the fundamental algorithms used in modern MILP solvers,and the selection of appropriate cuts from the candidate cuts subset is crucial for enhancing efficiency.Due to the reliance on expert knowledge and problem-specific heuristics,classical cut selection methods are not always transferable and often limit the scalability and generalizability of the cutting plane method.To provide a more efficient and generalizable strategy,we propose a reinforcement learning(RL)framework to enhance cut selection in the solving process of MILPs.Firstly,we design feature vectors to incorporate the inherent properties of MILP and computational information from the solver and represent MILP instances as bipartite graphs.Secondly,we choose the weighted metrics to approximate the proximity of feasible solutions to the convex hull and utilize the learning method to determine the weights assigned to each metric.Thirdly,a graph convolutional neural network is adopted with a self-attention mechanism to predict the value of weighting factors.Finally,we transform the cut selection process into a Markov decision process and utilize RL method to train the model.Extensive experiments are conducted based on a leading open-source MILP solver SCIP.Results on both general and specific datasets validate the effectiveness and efficiency of our proposed approach.

关键词： reinforcement learning mixed-integer linear programming cutting plane method cut selection

70 Gbps PAM-4850-nm oxide-confined VCSEL without equalization and pre-emphasis

在线全文

学校读者我要写书评

暂无评论

Journal of Semiconductors 2024年第5期45卷 5-7页

作者： Anjin Liu Bao Tang Zhiyong Li Wanhua Zheng Key Laboratory of Optoelectronic Materials and Devices Institute of SemiconductorsChinese Academy of SciencesBeijing 100083China Center of Materials Science and Optoelectronics Engineering University of Chinese Academy of SciencesBeijing 100049China Accelink technologies co. LtdWuhan 430205China Key Laboratory of Solid-State Optoelectronics information Technology Institute of SemiconductorsChinese Academy of SciencesBeijing 100083China

Directly modulated 850-nm vertical-cavity surface-emitting lasers(VCSELs)with the advantages of low cost,high modulation speed,good reliability,and low power consumption,are the key sources in the optical interconnects with multimode fibers for the supercomputers,data centers,and machine learning applications[1−3].Typically,non-return-tozero(NRZ)modulation format is used.

关键词： return confined fibers

A two-step multi-frequency receiver function inversion method for shallow crustal S-wave velocity structure and its application across the basin-mountain range belts in Northeast China

在线全文

学校读者我要写书评

暂无评论

Science China Earth sciences 2024年第3期67卷 687-703页

作者： Ruihao YANG Xu WANG Ling CHEN Mingye FENG Qifu CHEN State Key Laboratory of Lithospheric Evolution Institute of Geology and GeophysicsChinese Academy of SciencesBeijing 100029China College of Earth and Planetary sciences University of Chinese Academy of SciencesBeijing 100049China Key Laboratory of Earth and Planetary Physics Institute of Geology and GeophysicsChinese Academy of SciencesBeijing 100029China

A shallow crustal velocity structure(above 10 km depth) is essential for understanding the crustal structures and deformation and assessing the exploration prospect of natural resources, and also provides priori information for imaging deeper crustal and mantle structure. Passive-source seismic methods are cost-effective and advantageous for regional-scale imaging of shallow crustal structures compared to active-source methods. Among these passive methods, techniques utilizing receiver function waveforms and/or body-wave amplitude ratios have recently gained prominence due to their relatively high spatial resolution. However, in basin regions, reverberations caused by near-surface unconsolidated sedimentary layers often introduce strong non-uniqueness and uncertainty, limiting the applicability of such methods. To address these challenges, we propose a two-step inversion method that uses multi-frequency P-RF waveforms and P-RF horizontal-to-vertical amplitude ratios. Synthetic tests indicate that our two-step inversion method can mitigate the non-uniqueness of the inversion and enhance the stability of the results. Applying this method to teleseismic data from a linear seismic array across the sedimentary basins in Northeast China, we obtain a high-resolution image of the shallow crustal S-wave velocity structure along the array. Our results reveal significant differences between the basins and mountains. The identification of low-velocity anomalies(<2.8 km s^(-1)) at depths less than 1.0 km beneath the Erlian Basin and less than 2.5 km beneath the Songliao Basin suggests the existence of sedimentary layers. Moreover, the high-velocity anomalies(~3.4–3.8 km s^(-1)) occurring at depths greater than 7 km in the Songliao Basin may reflect mafic intrusions emplaced during the Early Cretaceous. Velocity anomaly distribution in our imaging result is consistent with the location of the major faults, uplifts, and sedimentary depressions, as well as active-source seismic results. This application further validates the eff

关键词： Receiver function Frequency dependence Two-step inversion Shallow crustal velocity structures Unconsolidated sedimentary basins

Angle measurement of pulsars based on spatially modulated X-ray intensity correlation

在线全文

学校读者我要写书评

暂无评论

Chinese Optics Letters 2024年第4期22卷 106-111页

作者：汪德喻虹谈志杰陆荣华韩申生 Key Laboratory of quantum Optics Shanghai Institute of Optics and Fine MechanicsChinese Academy of SciencesShanghai 201800China Center of Materials Science and Optoelectronics Engineering University of Chinese Academy of SciencesBeijing 100049China Hangzhou Institute for Advanced Study University of Chinese Academy of SciencesHangzhou 310024China

High-precision angle measurement of pulsars is critical for realizing pulsar navigation.Compared to visible light and radio waves,the wavelength of X-rays is incredibly short,which provides the possibility of achieving better spatial resolution.However,due to the lack of applicable X-ray apparatus,extracting the angle information of pulsars through conventional X-ray methods is challenging.Here,we propose an approach of pulsar angle measurement based on spatially modulated X-ray intensity correlation(SMXIC),in which the angle information is obtained by measuring the spatial intensity correlation between two radiation fields.The theoretical model for this method has been established,and a proof-of-concept experiment was carried out.The SMXIC measurement of observing angles has been demonstrated,and the experimental results are consistent with the theoretical values.The potential of this method in future applications is discussed,and theoretically,the angular measurement at the level of micro-arcsecond can be expected.The sphere of pulsar navigation may benefit from our fresh insights.

关键词： pulsar measurement X-ray measurement intensity correlation spatial modulation

A unified pre-training and adaptation framework for combinatorial optimization on graphs

在线全文

学校读者我要写书评

暂无评论

Science China Mathematics 2024年第6期67卷 1439-1456页

作者： Ruibin Zeng Minglong Lei Lingfeng Niu Lan Cheng School of Computer Science and Technology University of Chinese Academy of SciencesBeijing100049China Faculty of information Technology Beijing University of TechnologyBeijing100124China School of Economics and Management University of Chinese Academy of SciencesBeijing100190China School of Mathematics and Statistics Central South UniversityChangsha410075China

Combinatorial optimization(CO)on graphs is a classic topic that has been extensively studied across many scientific and industrial fields.Recently,solving CO problems on graphs through learning methods has attracted great attention.Advanced deep learning methods,e.g.,graph neural networks(GNNs),have been used to effectively assist the process of solving COs.However,current frameworks based on GNNs are mainly designed for certain CO problems,thereby failing to consider their transferable and generalizable abilities among different COs on graphs.Moreover,simply using original graphs to model COs only captures the direct correlations among objects,which does not consider the mathematical logicality and properties of COs.In this paper,we propose a unified pre-training and adaptation framework for COs on graphs with the help of the maximum satisfiability(Max-SAT)problem.We first use Max-SAT to bridge different COs on graphs since they can be converted to Max-SAT problems represented by standard formulas and clauses with logical information.Then we further design a pre-training and domain adaptation framework to extract the transferable and generalizable features so that different COs can benefit from them.In the pre-training stage,Max-SAT instances are generated to initialize the parameters of the model.In the fine-tuning stage,instances from CO and Max-SAT problems are used for adaptation so that the transferable ability can be further improved.Numerical experiments on several datasets show that features extracted by our framework exhibit superior transferability and Max-SAT can boost the ability to solve COs on graphs.

关键词： combinatorial optimization graph neural networks domain adaptation maximum satisfiability problem

Cloud Top Pressure Retrieval Using Polarized and Oxygen A-band Measurements from GF5 and PARASOL Satellites

在线全文

学校读者我要写书评

暂无评论

Advances in Atmospheric sciences 2024年第4期41卷 680-700页

作者： Lesi WEI Huazhe SHANG Jian XU Chong SHI Gegen TANA Kefu CHAO Shanhu BAO Liangfu CHEN Husi LETU State Key Laboratory of Remote Sensing Science Aerospace Information Research InstituteChinese Academy of SciencesBeijing 100101China National Space Science Center Chinese Academy of SciencesBeijing 101407China College of Physics and Electronic information Inner Mongolia Normal UniversityHohhot 010022China College of Geographical Science Inner Mongolia Normal UniversityHohhot 010022China University of Chinese academy of sciences Beijing 100101China

Cloud top pressure(CTP)is one of the critical cloud properties that significantly affects the radiative effect of clouds.Multi-angle polarized sensors can employ polarized bands(490 nm)or O_(2)A-bands(763 and 765 nm)to retrieve the CTP.However,the CTP retrieved by the two methods shows inconsistent results in certain cases,and large uncertainties in low and thin cloud retrievals,which may lead to challenges in subsequent applications.This study proposes a synergistic algorithm that considers both O_(2)A-bands and polarized bands using a random forest(RF)model.LiDAR CTP data are used as the true values and the polarized and non-polarized measurements are concatenated to train the RF model to determine CTP.Additionally,through analysis,we proposed that the polarized signal becomes saturated as the cloud optical thickness(COT)increases,necessitating a particular treatment for cases where COT<10 to improve the algorithm's stability.The synergistic method was then applied to the directional polarized camera(DPC)and Polarized and Directionality of the Earth’s Reflectance(POLDER)measurements for evaluation,and the resulting retrieval accuracy of the POLDER-based measurements(RMSEPOLDER=205.176 hPa,RMSEDPC=171.141 hPa,R^(2)POLDER=0.636,R^(2)DPC=0.663,respectively)were higher than that of the MODIS and POLDER Rayleigh pressure measurements.The synergistic algorithm also showed good performance with the application of DPC data.This algorithm is expected to provide data support for atmosphere-related fields as an atmospheric remote sensing algorithm within the Cloud Application for Remote Sensing,Atmospheric Radiation,and Updating Energy(CARE)platform.

关键词： atmospheric remote sensing cloud top pressure multi-angle polarized O_(2)A-band GF-5/DPC PARASOL/POLDER-3

Remote entangling gate between a quantum dot spin and a transmon qubit mediated by microwave photons

在线全文

学校读者我要写书评

暂无评论

Chinese Physics B 2024年第2期33卷 52-59页

作者：朱行宇朱乐天涂涛李传锋 Key Laboratory of quantum information Chinese Academy of SciencesUniversity of Science and Technology of ChinaHefei 230026China School of Mechanical and Electronic Engineering Suzhou UniversitySuzhou 234000China Hefei National Laboratory University of Science and Technology of ChinaChinese Academy of SciencesHefei 230088China

Spin qubits and superconducting qubits are promising candidates for realizing solid-state quantum information processors.Designing a hybrid architecture that combines the advantages of different qubits on the same chip is a highly desirable but challenging goal.Here we propose a hybrid architecture that utilizes a high-impedance SQUID array resonator as a quantum bus,thereby coherently coupling different solid-state qubits.We employ a resonant exchange spin qubit hosted in a triple quantum dot and a superconducting transmon qubit.Since this hybrid system is highly tunable,it can operate in a dispersive regime,where the interaction between the different qubits is mediated by virtual photons.By utilizing such interactions,entangling gate operations between different qubits can be realized in a short time of 30 ns with a fidelity of up to 96.5%under realistic parameter conditions.Further utilizing this interaction,remote entangled state between different qubits can be prepared and is robust to perturbations of various parameters.These results pave the way for exploring efficient fault-tolerant quantum computation on hybrid quantum architecture platforms.

关键词： hybrid quantum architectures circuit quantum electrodynamics entangling gate

Refined symmetry-resolved Page curve and charged black holes

在线全文

学校读者我要写书评

暂无评论

Chinese Physics C 2024年第5期48卷 134-140页

作者：李磐凌意 Institute of High Energy Physics Chinese Academy of SciencesBeijing 100049China School of Physics University of Chinese Academy of SciencesBeijing 100049China

The Page curve plotted using the typical random state approximation is not applicable to a system with conserved quantities,such as the evaporation process of a charged black hole,during which the electric charge does not macroscopically radiate out with a uniform rate.In this context,the symmetry-resolved entanglement entropy may play a significant role in describing the entanglement structure of such a system.We attempt to impose constraints on microscopic quantum states to match the macroscopic phenomenon of charge radiation during black hole evaporation.Specifically,we consider a simple qubit system with conserved spin/charge serving as a toy model for the evaporation of charged black holes.We propose refined rules for selecting a random state with conserved quantities to simulate the distribution of charges during the different stages of evaporation and obtain refined Page curves that exhibit distinct features in contrast to the original Page curve.We find that the refined Page curve may have a different Page time and exhibit asymmetric behavior on both sides of the Page time.Such refined Page curves may provide a more realistic description for the entanglement between the charged black hole and radiation during the evaporation process.

关键词： quantum entanglement Page curve charged black hole black hole evaporation Hawking radiation Schwinger effect random state average entropy

NFT Security Matrix:Towards Modeling NFT Ecosystem Threat

在线全文

学校读者我要写书评

暂无评论

Computer Modeling in Engineering & sciences 2024年第6期139卷 3255-3285页

作者： Peng Liao Chaoge Liu Jie Yin Zhi Wang Xiang Cui Key Laboratory of Trustworthy Distributed Computing and Service Beijing University of Posts and TelecommunicationsBeijing100876China Zhongguancun Laboratory Beijing100094China Institute of information Engineering Chinese Academy of SciencesBeijing100085China

Digital assets have boomed over the past few years with the emergence of Non-fungible Tokens(NFTs).To be specific,the total trading volume of digital assets reached an astounding$55.5 billion in 2022.Nevertheless,numerous security concerns have been raised by the rapid expansion of the NFT ecosystem.NFT holders are exposed to a plethora of scams and traps,putting their digital assets at risk of being lost.However,academic research on NFT security is scarce,and the security issues have aroused rare attention.In this study,the NFT ecological process is comprehensively explored.This process falls into five different stages encompassing the entire lifecycle of NFTs.Subsequently,the security issues regarding the respective stage are elaborated and analyzed in depth.A matrix model is proposed as a novel contribution to the categorization of NFT security issues.Diverse data are collected from social networks,the Ethereum blockchain,and NFT markets to substantiate our claims regarding the severity of security concerns in the NFT ecosystem.From this comprehensive dataset,nine key NFT security issues are identified from the matrix model and then subjected to qualitative and quantitative analysis.This study aims to shed light on the severity of NFT ecosystem security issues.The findings stress the need for increased attention and proactive measures to safeguard the NFT ecosystem.

关键词： Non-fungible token blockchain cyber security