检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Wang, Zhuoyuan Mi, Jiacong Lu, Shan He, Jieyue School of Computer Science and Engineering Key Lab of Computer Network and Information Integration MOE Southeast University Jiangsu Nanjing210018 China Nanjing Fenghuo Tiandi Communication Technology Co. Ltd. Jiangsu Nanjing211161 China

The quest for accurate prediction of drug molecule properties poses a fundamental challenge in the realm of Artificial Intelligence Drug Discovery (AIDD). An effective representation of drug molecules emerges as a pivotal component in this pursuit. Contemporary leading-edge research predominantly resorts to self-supervised learning (SSL) techniques to extract meaningful structural representations from large-scale, unlabeled molecular data, subsequently fine-tuning these representations for an array of downstream tasks. However, an inherent shortcoming of these studies lies in their singular reliance on one modality of molecular information, such as molecule image or SMILES representations, thus neglecting the potential complementarity of various molecular modalities. In response to this limitation, we propose MolIG, a novel MultiModaL molecular pre-training framework for predicting molecular properties based on Image and Graph structures. MolIG model innovatively leverages the coherence and correlation between molecule graph and molecule image to execute self-supervised tasks, effectively amalgamating the strengths of both molecular representation forms. This holistic approach allows for the capture of pivotal molecular structural characteristics and high-level semantic information. Upon completion of pre-training, Graph Neural Network (GNN) Encoder is used for the prediction of downstream tasks. In comparison to advanced baseline models, MolIG exhibits enhanced performance in downstream tasks pertaining to molecular property prediction within benchmark groups such as MoleculeNet Benchmark Group and ADMET Benchmark Group. © 2023, CC BY.

关键词： Molecules

来源：评论

学校读者我要写书评

暂无评论

FINDER: A Simple and Effective Defender against Unnoticeable Graph Injection Attacks

FINDER: A Simple and Effective Defender against Unnoticeable...

引用

IEEE International Conference on Trust, Security and Privacy in Computing and Communications (TrustCom)

作者： Linlin Su Jinyan Wang Zeming Gan Xianxian Li Key Lab of Education Blockchain and Intelligent Technology Ministry of Education Guangxi Normal University Guilin China Guangxi Key Lab of Multi-Source Information Mining and Security Guangxi Normal University Guilin China School of Computer Science and Engineering Guangxi Normal University Guilin China

ISBN: (数字)9798350381993

ISBN: (纸本)9798350382006

Graph Neural Networks (GNNs) have various applications in real-life scenarios. However, they are vulnerable to adversarial attacks, especially Graph Injection Attacks (GIAs). The flexible and high-risk GIAs pose a significant threat by injecting malicious nodes into the graph. Regrettably, defense strategies to resist GIAs are still scarce. The current defenders lack comprehensive strategies that can effectively resist a wide range of injection techniques. Particularly, when confronted with highly unnoticeable GIAs, the performance of these defenses is significantly diminished. In this paper, we propose a simple and effective GIA defense method named Faithful INjection DefendER (FINDER). FINDER evaluates the impact of the nodes on their neighbors in the context of message aggregation to identify the potential injected nodes in the graph. By blocking identified injected nodes from interacting with the benign nodes, FINDER prevents the further propagation of misinformation and the spread of adversarial influence throughout the graph. Through extensive experiments conducted on three benchmarks, FINDER demonstrates remarkable performance in accurately identifying injected nodes and defending various GIAs.

关键词： Privacy Resists Benchmark testing Graph neural networks Security Fake news

来源：评论

学校读者我要写书评

暂无评论

Flow Simulation of Lid-Driven Rectangular Cavity by Using Lattice Boltzmann Method

arXiv

引用

arXiv 2022年

作者： Xiang, Xiuqiao Shi, Baochang School of Computer Science China University of Geosciences Wuhan430078 China School of Mathematics and Statistics Huazhong University of Science and Technology Wuhan430074 China National Engineering Research Center of Geographic Information System China University of Geosciences Wuhan430078 China

Wall-driven flow in square cavity has been studied extensively, yet it is more frequently for the rectangular cavity flow occurring in practical problems, and some flow characteristics about rectangular cavity have not been fully investigated. As a promising numerical simulation tool, the Lattice Boltzmann Method (LBM) is employed to simulate the lid-driven flow in a two-dimensional rectangular cavity in this paper. First, the code is validated for the standard square cavity, the velocity profiles, stream function values and center positions of the primary and second vortexes at different Re are presented and compared with previous researches. Then, the eddy dynamics of rectangular cavities is simulated and discussed with Reynolds number (Re) in the range of 4000-8000 and vertical to horizontal axis ratio (Ar) varied from 0.4 to 2.0, and the streamline and center migration of the primary vortex are drawn realistically for rectangular cavity. In the end, the steady, periodic, aperiodic and unstable phenomenon in the rectangular cavity is produced by LBM, Re and Ar which play the significant role in the state transition of lid-driven rectangular cavity flow are analyzed and summarized in detail. By the LBM numerical simulation with 500×(500*Ar) grid, we have discovered that the flow state in the rectangular cavity with a fixed Ar changes monotonically from stable state, periodic state to aperiodic and unstable state with the increase of Re, while the flow state in the rectangular cavity with a fixed Re varies non-monotonically with the increase of Ar. Additionally, the cycle length of periodic states varies with the change of Ar and Re. Copyright © 2022, The Authors. All rights reserved.

关键词： Reynolds number

来源：评论

学校读者我要写书评

暂无评论

Unsupervised Pre-training with Language-Vision Prompts for Low-Data Instance Segmentation

arXiv

引用

arXiv 2024年

作者： Zhang, Dingwen Li, Hao He, Diqi Liu, Nian Cheng, Lechao Wang, Jingdong Han, Junwei The Brain and Artificial Intelligence Lab. Northwestern Polytechnical University Shaanxi Xi’an710000 China The Inception Institute of Artificial Intelligence United Kingdom The School of Computer Science and Information Engineering Hefei University of Technology Hefei230601 China Department of Computer Vision Baidu Inc. China

In recent times, following the paradigm of DETR (DEtection TRansformer), query-based end-to-end instance segmentation (QEIS) methods have exhibited superior performance compared to CNN-based models, particularly when trained on large-scale datasets. Nevertheless, the effectiveness of these QEIS methods diminishes significantly when confronted with limited training data. This limitation arises from their reliance on substantial data volumes to effectively train the pivotal queries/kernels that are essential for acquiring localization and shape priors. To address this problem, we propose a novel method for unsupervised pre-training in low-data regimes. Inspired by the recently successful prompting technique, we introduce a new method, Unsupervised Pre-training with Language-Vision Prompts (UPLVP), which improves QEIS models’ instance segmentation by bringing language-vision prompts to queries/kernels. Our method consists of three parts: (1) Masks Proposal: Utilizes language-vision models to generate pseudo masks based on unlabeled images. (2) Prompt-Kernel Matching: Converts pseudo masks into prompts and injects the best-matched localization and shape features to their corresponding kernels. (3) Kernel Supervision: Formulates supervision for pre-training at the kernel level to ensure robust learning. With the help of our pre-training method, QEIS models can converge faster and perform better than CNN-based models in low-data regimes. Experimental evaluations conducted on MS COCO, Cityscapes, and CTW1500 datasets indicate that the QEIS models’ performance can be significantly improved when pre-trained with our method. Code will be available at: https://***/lifuguan/UPLVP. Copyright © 2024, The Authors. All rights reserved.

关键词： Large datasets

来源：评论

学校读者我要写书评

暂无评论

Enhanced THz Emission and Chirality Control in van der Waals Ferromagnetic FePd2Te2/Pt Heterostructures

引用

Journal of the American Chemical Society 2025年第23期147卷 2025 May 27页

作者： Jiali Zhang Bingxian Shi Haoran Xu Yiwen Song Yuqing Zou Ziyang Li Hongtao Dai Yuna Song Qingyuan Jin Peng Cheng Zuanming Jin Zongzhi Zhang Key Laboratory of Micro and Nano Photonic Structures (MOE) School of Information Science and Technology Fudan University Shanghai 200433 China. Department of Physics Key Laboratory of Quantum State Construction and Manipulation Ministry of Education Renmin University of China Beijing 100872 China. Laboratory for Neutron Scattering Department of Physics Renmin University of China Beijing 100872 China. Terahertz Technology Innovation Research Institute Terahertz Spectrum and Imaging Technology Cooperative Innovation Center Shanghai Key Lab of Modern Optical System University of Shanghai for Science and Technology Shanghai 200093 China. State Key Laboratory of Precision Spectroscopy East China Normal University Shanghai 200062 P. R. China.

Two-dimensional (2D) magnetic materials, with their unique van der Waals (vdW) layered structure, tunable magnetism, and strong spin-orbit coupling, are promising for spintronic applications like terahertz (THz) emitters. Our study investigates the THz radiation properties of the newly discovered vdW ferromagnet FePdTe, which features a one-dimensional Fe zigzag chain structure and strong in-plane uniaxial anisotropy. Using ultrafast reflective THz emission spectroscopy (TES), we demonstrate that the paramagnetic FePdTe produces weak THz emission at room temperature. However, when capped with a 3 nm Pt layer, the THz emission intensity is significantly enhanced, attributed to spin-to-charge conversion (SCC) via the inverse spin Hall effect (ISHE). In the FePdTe/Pt heterostructure, the THz emission intensity is closely linked to the spin texture induced by crystal twinning in FePdTe and depends on the polarization angle of the pump laser. Two mechanisms for the photoexcited spin current generation are identified: one from spin-polarized hot electrons at the FePdTe/Pt interface, and the other, which is dominant, from angular momentum transfer through a nonlinear optical effect that induces impulsive magnetization. These findings provide valuable insights into the spin dynamics of 2D materials and open avenues for high-density, low-power THz spintronic device development based on advanced 2D vdW magnets.

关键词： Van der Waals forces

来源：评论

学校读者我要写书评

暂无评论

Smartsense: Efficient Snapshotting in Eh-Wsns

SSRN

引用

SSRN 2023年

作者： Singla, Priyanka Sarangi, Smruti R. The School of Information Technology IIT Delhi New Delhi India Lab No. 306 SIT Department IIT Delhi Hauz Khas New Delhi110016 India The Department of Computer Science and Engineering IIT Delhi New Delhi India

Energy harvesting wireless sensor networks (EH-WSNs) play a vital role in environmental monitoring. The sensor nodes sense their ambience and transmit data to a remote sink node via multi-hop communication, enabling informed decision making at the sink. To make accurate decisions, it’s crucial to have information about monitored parameters across the entire network simultaneously, i.e., capturing a network snapshot. However, the variability in ambient energy supply hampers successful communication among these nodes, making snapshotting difficult and rendering conventional WSN-based solutions ineffective. Moreover, network congestion and varying distances between individual nodes and the sink node result in messages not reaching the sink node simultaneously, further complicating the snapshot process. Unfortunately, state-of-the-art algorithms produce many ineffectual messages that do not enhance snapshot accuracy;instead, they contribute to increased network traffic. Thus, there is a pressing need for a tailor-made algorithm designed for snapshotting in EH-WSNs. This algorithm should be capable of capturing a large number of snapshots. In this direction, we propose SmartSense– a novel distributed system in which nodes adapt their sensing periods and transmission times based on their physical locations, available ambient energy, and network congestion. We show SmartSense’s efficacy by comparing it against four different systems: two theoretical baselines and two state-of-the-art-systems. The two theoretical baselines consist of hypothetical omniscient algorithms: (i) an AI-based solution that dynamically changes the sensing periods in response to the ambient energy available at nodes and (ii) a static solution that comprises the best possible sensing periods after experimenting with a huge number of sensing periods a priori. Our simulations show that SmartSense’s effective throughput is within [90-101.8]% of the dynamic AI-based solution and [1.8-75]% better than the

关键词： Energy harvesting

来源：评论

学校读者我要写书评

暂无评论

Breakup dynamics of dilute polymer solution droplet moving down a grooved hydrophobic surface

引用

Physics of Fluids 2025年第5期37卷

作者： Cho, Yujin Kim, Uijin Kim, Jeong-Hyun Department of Mechanical Information Engineering Seoul National University of Science and Technology Seoul01811 Korea Republic of Department of Mechanical System Design Engineering Seoul National University of Science and Technology Seoul01811 Korea Republic of

In this study, we investigated the behavior of aqueous polymer solution droplets moving down an inclined, grooved hydrophobic surface using a shadow imaging system. Several aqueous polyethylene oxide (PEO) solutions were prepared by adding different amounts of PEO with molecular weights of Mw = 1 × 106 g/mol in water. Rectangular grooves were fabricated on the polydimethylsiloxane (PDMS) surface using standard lithography. The receding dynamics of droplets at various surface-inclination angles were captured using high-speed imaging. The increase in viscosity and elasticity of the droplets significantly delayed the retraction of the liquid on top of the grooves, leading to the formation of several liquid films between the grooves. These films, which resembled planar surfaces, evolved into cylindrical filaments, with the filament diameter decreasing exponentially over time. As the polymer concentration and surface-inclination angle increased, the lifetime of the liquid ligaments increased owing to the increased elasticity of the liquid. The extensional relaxation times of all the dilute polymer solution droplets used in the experiment were calculated by evaluating the slope of the filament decay in the elasto-capillary regime, and they closely matched the values reported in the literature. Our findings not only elucidate the receding dynamics of dilute polymer solution droplets but also open up the possibility of using the sliding droplet method (SDM) to quantify the extensional properties of these solutions. © 2025 Author(s).

关键词： High speed photography

来源：评论

学校读者我要写书评

暂无评论

H∞$$ {H}_{\infty } $$PID Control for Singularly Perturbed systems With Randomly Switching Nonlinearities Under Dynamic Event-Triggered Mechanism

引用

International Journal of Robust and Nonlinear Control 2025年

作者： Zhenglong Jiao Zidong Wang Yuqiang Luo Fan Wang Shanghai Key Lab of Modern Optical System Department of Control Science and Engineering University of Shanghai for Science and Technology Shanghai China Department of Computer Science Brunel University London Uxbridge Middlesex UK School of Automation Nanjing University of Information Science and Technology Nanjing China

In this article, the observer-based H ∞ $$ {H}_{\infty } $$ proportional-integral-derivative (PID) control problem is studied for a class of singular perturbed systems with randomly switching nonlinearities. During the modeling process, a set of binary random sequences is introduced to describe the random behaviors of the nonlinear switching encountered in engineering practice. Considering the singular perturbation parameter (SPP), an observer-based H ∞ $$ {H}_{\infty } $$ PID controller is constructed for the concerned singularly perturbed systems (SPSs), where the state information is estimated. To improve communication efficiency and reduce resource waste, a dynamic event-triggered mechanism is employed during the measurement transmission. Meanwhile, an SPP-dependent Lyapunov-Krasovskii function is utilized to derive conditions under which the closed-loop system is guaranteed to be stochastically stable with the desired H ∞ $$ {H}_{\infty } $$ performance under a specified upper bound on the SPP. All the desired observer and controller gains are determined by solving a set of matrix inequalities. Finally, the effectiveness and superiority of the suggested PID control method are verified through a simulation example.

关键词： H∞$$ {H}_{\infty } $$ performance index dynamic event-triggered mechanism proportional-integral-derivative control randomly switching nonlinearities singularly perturbed systems

来源：评论

学校读者我要写书评

暂无评论

Variational quantum eigensolver with linear depth problem-inspired ansatz for solving portfolio optimization in finance

arXiv

引用

arXiv 2024年

作者： Wang, Shengbin Wang, Peng Li, Guihui Zhao, Shubin Zhao, Dongyi Wang, Jing Fang, Yuan Dou, Menghan Gu, Yongjian Wu, Yu-Chun Guo, Guo-Ping Origin Quantum Computing Company Limited Anhui Hefei230026 China CAS Key Laboratory of Quantum Information University of Science and Technology of China Anhui Hefei230026 China Intelligent Information Sensing and Processing Lab College of Electronic Engineering Ocean University of China Shandong Qingdao266000 China College of Physics and Optoelectronic Engineering Ocean University of China Shandong Qingdao266100 China Institute of Artificial Intelligence Hefei Comprehensive National Science Center Anhui Hefei230088 China

Great efforts have been dedicated in recent years to explore practical applications for noisy intermediate-scale quantum (NISQ) computers, which is a fundamental and challenging problem in quantum computing. As one of the most promising methods, the variational quantum eigensolver (VQE) has been extensively studied. In this paper, VQE is applied to solve portfolio optimization problems in finance by designing two hardware-efficient Dicke state ansatze that reach a maximum of 2n two-qubit 2 gate depth and n2/4 parameters, with n being the number of qubits used. Both ansatze are partitioning-friendly, allowing for the proposal of a highly scalable quantum/classical hybrid distributed computing (HDC) scheme. Combining simultaneous sampling, problem-specific measurement error mitigation, and fragment reuse techniques, we successfully implement the HDC experiments on a superconducting quantum computer with up to 55 qubits. The simulation and experimental results illustrate that the restricted expressibility of the ansatze, induced by the small number of parameters and limited entanglement, is advantageous for solving classical optimization problems with the cost function of the conditional value-at-risk (CVaR) for the NISQ era and beyond. Furthermore, the HDC scheme shows great potential for achieving quantum advantage in the NISQ era. We hope that the heuristic idea presented in this paper can motivate fruitful investigations in current and future quantum computing paradigms. Copyright © 2024, The Authors. All rights reserved.

关键词： Optimization

来源：评论

学校读者我要写书评

暂无评论

AgileWatts: An Energy-Efficient CPU Core Idle-State Architecture for Latency-Sensitive Server Applications 22

AgileWatts: An Energy-Efficient CPU Core Idle-State Architec...

引用

Proceedings of the 55th Annual IEEE/ACM International Symposium on Microarchitecture

作者： Jawad Haj Yahya Haris Volos Davide B. Bartolini Georgia Antoniou Jeremie S. Kim Zhe Wang Kleovoulos Kalaitzidis Tom Rollet Zhirui Chen Ye Geng Onur Mutlu Yiannakis Sazeides Rivos Inc Mountain View California United States Computer Architecture Laboratory Department of Computer Science University of Cyprus Computing Systems Lab Huawei Zurich Research Center Department of Information Technology and Electrical Engineering ETH Zurich Switzerland

ISBN: (纸本)9781665462723

User-facing applications running in modern datacenters exhibit irregular request patterns and are implemented using a multitude of services with tight latency requirements (30--250μs). These characteristics render existing energy-conserving techniques ineffective when processors are idle due to the long transition time (order of 100μs) from a deep CPU core idle power state (C-state). While prior works propose management techniques to mitigate this inefficiency, we tackle it at its root with AgileWatts (AW): a new deep CPU core C-state architecture optimized for datacenter server processors targeting latency-sensitive *** drastically reduces the transition latency from deep CPU core idle power states while retaining most of their power savings based on three key ideas. First, AW eliminates the latency (several microseconds) of saving/restoring the core context when powering-off/-on the core in a deep idle state by i) implementing medium-grained power-gates, carefully distributed across the CPU core, and ii) retaining context in the power-ungated domain. Second, AW eliminates the flush latency (several tens of microseconds) of the L1/L2 caches when entering a deep idle state by keeping L1/L2 content power-ungated. A small control logic also remains ungated to serve cache coherence traffic. AW implements cache sleep-mode and leakage reduction for the power-ungated domain by lowering a core's voltage to the minimum operational level. Third, using a state-of-the-art power efficient all-digital phase-locked loop (ADPLL) clock generator, AW keeps the PLL active and locked during the idle state, cutting microseconds of wake-up latency at negligible power *** evaluation with an accurate industrial-grade simulator calibrated against an Intel Skylake server shows that AW reduces the energy consumption of Memcached by up to 71% (35% on average) with <1% end-to-end performance degradation. We observe similar trends for other evaluated services (MySQL and Kafka).

关键词：

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：