检索结果-内蒙古大学图书馆

CiRLExplainer: Causality-Inspired Explainer for Graph Neural Networks via Reinforcement Learning

IEEE Transactions on Neural Networks and Learning Systems 2025年第6期36卷 9970-9984页

作者： Wenya Hu Jia Wu Quan Qian School of Computer Engineering and Science Shanghai University Shanghai China School of Computing Macquarie University Sydney NSW Australia School of Computer Engineering and Science the Center of Materials Informatics and Data Science Materials Genome Institute the Key Laboratory of Silicate Cultural Relics Conservation Ministry of Education and Shanghai Institute for Advanced Communication and Data Science Shanghai University Shanghai China

In this article, we propose a new graph neural network (GNN) explainability model, CiRLExplainer, which elucidates GNN predictions from a causal attribution perspective. Initially, a causal graph is constructed to analyze the causal relationships between the graph structure and GNN predicted values, identifying node attributes as confounding factors between the two. Subsequently, a backdoor adjustment strategy is employed to circumvent these confounders. Additionally, since the edges within the graph structure are not independent, reinforcement learning is incorporated. Through a sequential selection process, each step evaluates the combined effects of an edge and the previous structure to generate an explanatory subgraph. Specifically, a policy network predicts the probability of each candidate edge being selected and adds a new edge through sampling. The causal effect of this action is quantified as a reward, reflecting the interactivity among edges. By maximizing the policy gradient during training, the reward stream of the edge sequence is optimized. The CiRLExplainer is versatile and can be applied to any GNN model. A series of experiments was conducted, including accuracy (ACC) analysis of the explanation results, visualization of the explanatory subgraph, and ablation studies considering node attributes as confounding factors. The experimental results demonstrate that our model not only outperforms current state-of-the-art explanation techniques, but also provides precise semantic explanations from a causal perspective. Additionally, the experiments validate the rationale for considering node attributes as confounding factors, thereby enhancing the explanatory power and ACC of the model. Notably, across different datasets, our explainer achieved improvements over the best baseline models in the ACC-area under the curve (AUC) metrics by 5.89%, 5.69%, and 4.87%, respectively.

关键词： Predictive models Graph neural networks Reinforcement learning Accuracy Training Prediction algorithms Analytical models Sustainable development Focusing data science

来源：评论

学校读者我要写书评

暂无评论

Generation of longitudinally polarized multi-segment optical needles by tightly focusing RPBG beam

引用

Optoelectronics Letters 2023年第7期19卷 399-404页

作者： SHI Changkun SONG Yiding DONG Bing ZHOU Zhanqi ZHANG Zengqi XU Zongwei State Key Laboratory of Precision Measuring Technology&Instruments Tianjin UniversityTianjin 300072China School of Optical-Electrical and Computer Engineering University of Shanghai for Science and TechnologyShanghai 200093China

In this paper,an effective method to garner sub-wavelength longitudinally polarized multi-segment optical needle sequence by using a specially designed hybrid filter(HF) in a high numerical aperture(NA) objective focusing system is *** HF is coupled by a binary phase transmission function and a multi-segment modulation function,and the binary phase filter is designed by the particle swarm optimization(PSO) algorithm and acts on the radially polarized Bessel Gaussian(RPBG) beam to obtain a longitudinally polarized optical needle with long depth of focus(DOF,6λ) and a sub-wavelength transverse spot size(0.430λ).The optical needle is with high uniformity of 98% and high beam quality of 96%,and the negligible sidelobe is 15%.On this basis,the multi-segment optical needle sequence with tunable spacing or number can be realized by the multi-segment modulation *** is found that the HF makes the generation of multi-segment optical needle sequence more flexible and *** research has broad application prospects in material processing,particle acceleration,particle capture and other fields.

关键词： function polarized beam

来源：评论

学校读者我要写书评

暂无评论

PGN: The RNN's New Successor is Effective for Long-Range Time Series Forecasting 38

PGN: The RNN's New Successor is Effective for Long-Range Tim...

引用

38th Conference on Neural Information Processing Systems, NeurIPS 2024

作者： Jia, Yuxin Lin, Youfang Yu, Jing Wang, Shuo Liu, Tianhao Wan, Huaiyu School of Computer Science & Technology Beijing Jiaotong University China Beijing Key Laboratory of Traffic Data Analysis and Mining Beijing China School of Cyberspace Science and Technology Beijing Jiaotong University China

Due to the recurrent structure of RNN, the long information propagation path poses limitations in capturing long-term dependencies, gradient explosion/vanishing issues, and inefficient sequential execution. Based on this, we propose a novel paradigm called Parallel Gated Network (PGN) as the new successor to RNN. PGN directly captures information from previous time steps through the designed Historical Information Extraction (HIE) layer and leverages gated mechanisms to select and fuse it with the current time step information. This reduces the information propagation path to O(1), effectively addressing the limitations of RNN. To enhance PGN's performance in long-range time series forecasting tasks, we propose a novel temporal modeling framework called Temporal PGN (TPGN). TPGN incorporates two branches to comprehensively capture the semantic information of time series. One branch utilizes PGN to capture long-term periodic patterns while preserving their local characteristics. The other branch employs patches to capture short-term information and aggregate the global representation of the series. TPGN achieves a theoretical complexity of O(√L), ensuring efficiency in its operations. Experimental results on five benchmark datasets demonstrate the state-of-the-art (SOTA) performance and high efficiency of TPGN, further confirming the effectiveness of PGN as the new successor to RNN in long-range time series forecasting. The code is available in this repository: https://***/Water2sea/TPGN. © 2024 Neural information processing systems foundation. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Lithographic printing inspired in-situ transfer of MXene-based films with localized topo-electro tunability for high-performance flexible pressure sensors

引用

Nano Research 2023年第11期16卷 12670-12679页

作者： Qiuyang Yan Yi Zhou Yin Cheng Liangjing Shi Ranran Wang Lian Gao Jing Sun State Key Laboratory of High Performance Ceramics and Superfine Microstructure Shanghai Institute of CeramicsChinese Academy of ScienceShanghai 200050China Center of Materials Science and Optoelectronics Engineering University of Chinese Academy of SciencesBeijing 100049China Department of Electrical and Computer Engineering National University of SingaporeSingapore 117583Singapore School of Chemistry and Materials Science Hangzhou Institute for Advanced StudyUniversity of Chinese Academy of SciencesHangzhou 310024China School of Materials Science and Engineering Shanghai Jiao Tong UniversityShanghai 200240China

MXene-based films have been intensively explored for construction of piezoresistive flexible pressure sensors owing to their excellent mechanical and electrical *** pressure sensitivity relies on pre-molding a flexible substrate,or regulating the micromorphology of MXene sheets,to obtain a micro-structured ***,the two avenues usually require complicated and time-consuming microfabrication or wet chemical processing,and are limited to non-adjustable topographicelectrical(topo-electro)***,we propose a lithographic printing inspired in-situ transfer(LIPIT)strategy to fabricate MXene-ink films(MIFs).In LIPIT,MIFs not only inherit ridge-and-valley microstructure from paper substrate,but also achieve localized topo-electro tunability by programming ink-writing patterns and *** MIF-based flexible pressure sensor with periodical topo-electro gradient exhibits remarkably boosted sensitivity in a wide sensing range(low detection limit of 0.29 Pa and working range of 100 kPa).The MIF sensor demonstrates versatile applicability in both subtle and vigorous pressuresensing fields,ranging from pulse wave extraction and machine learning-assisted surface texture recognition to piano-training glove(PT-glove)for piano *** LIPIT is quick,low-cost,and compatible with free ink/substrate combinations,which promises a versatile toolbox for designing functional MXene films with tailored morphological-mechanical-electrical properties for extended application scenarios.

关键词： MXene film transfer printing surface tunability flexible pressure sensor

来源：评论

学校读者我要写书评

暂无评论

Team Formation and Task Recommendation for Disabled People in Crowdsourcing Systems 25

Team Formation and Task Recommendation for Disabled People i...

引用

25th IEEE International Conference on computer Supported Cooperative Work in Design, CSCWD 2022

作者： Liang, Yifan Zou, Zongzheng Zhang, Peng Li, Dongsheng Lu, Tun Gu, Ning School of Computer Science Fudan University Shanghai China Shanghai Key Laboratory of Data Science Fudan University Shanghai China Microsoft Research Asia Shanghai China

ISBN: (纸本)9781665405270

The barrier-free map is a new type of map, the demand for which is increasing in modern society, and it is necessary to rely on crowdsourcing platforms to complete the labeling work of barrier-free facilities. Nowadays more people with disabilities are participating in crowdsourcing work. They tend to work with familiar people and hope to maximize their value. However, there is very little research about automatic management of organizations and tasks for the disabled to assist them to complete the work on the crowdsourcing platform. In this paper, we design a new team formation method based on the theory of socially embedded work of the disabled (TFSEW) and a bi-objective task recommendation method (BOTR), in order to ensure the efficiency of task completion and promote the participation of the disabled, thereby realizing their value in the work environment. Our method proves to be superior through experiments. At the same time, through questionnaire, people with disabilities show great anticipation for participating in the tasks in form of team and receive tasks recommended by the system. © 2022 IEEE.

关键词： Crowdsourcing

来源：评论

学校读者我要写书评

暂无评论

Head-wise Shareable Attention for Large Language Models

Head-wise Shareable Attention for Large Language Models

引用

2024 Conference on Empirical Methods in Natural Language Processing, EMNLP 2024

作者： Cao, Zouying Yang, Yifei Zhao, Hai Department of Computer Science and Engineering Shanghai Jiao Tong University China Key Laboratory of Shanghai Education Commission for Intelligent Interaction and Cognitive Engineering Shanghai Jiao Tong University China Shanghai Key Laboratory of Trusted Data Circulation and Governance in Web3 China

ISBN: (纸本)9798891761681

Large Language Models (LLMs) suffer from huge number of parameters, which restricts their deployment on edge devices. Weight sharing is one promising solution that encourages weight reuse, effectively reducing memory usage with less performance drop. However, current weight sharing techniques primarily focus on small-scale models like BERT and employ coarse-grained sharing rules, e.g., layer-wise. This becomes limiting given the prevalence of LLMs and sharing an entire layer or block obviously diminishes the flexibility of weight sharing. In this paper, we present a perspective on head-wise shareable attention for large language models. We further propose two memory-efficient methods that share parameters across attention heads, with a specific focus on LLMs. Both of them use the same dynamic strategy to select the shared weight matrices. The first method directly reuses the pre-trained weights without retraining, denoted as DirectShare. The second method first post-trains with constraint on weight matrix similarity and then shares, denoted as PostShare. Experimental results reveal our head-wise shared models still maintain satisfactory capabilities, demonstrating the feasibility of fine-grained weight sharing applied to LLMs. © 2024 Association for Computational Linguistics.

关键词： Computational linguistics

来源：评论

学校读者我要写书评

暂无评论

Calibrating the Confidence of Large Language Models by Eliciting Fidelity

Calibrating the Confidence of Large Language Models by Elici...

引用

2024 Conference on Empirical Methods in Natural Language Processing, EMNLP 2024

作者： Zhang, Mozhi Huang, Mianqiu Shi, Rundong Guo, Linsen Peng, Chong Yan, Peng Zhou, Yaqian Qiu, Xipeng School of Computer Science Fudan University China Shanghai Key Laboratory of Intelligent Information Processing Fudan University China Meituan China

ISBN: (纸本)9798891761643

Large language models optimized with techniques like RLHF have achieved good alignment in being helpful and harmless. However, post-alignment, these language models often exhibit overconfidence, where the expressed confidence does not accurately calibrate with their correctness rate. In this paper, we decompose the language model confidence into the Uncertainty about the question and the Fidelity to the answer generated by language models. Then, we propose a plug-and-play method, UF Calibration, to estimate the confidence of language models. Our method has shown good calibration performance by conducting experiments with 6 RLHF-LMs on four MCQA datasets. Moreover, we propose two novel metrics, IPR and CE, to evaluate the calibration of the model, and we have conducted a detailed discussion on Truly Well-Calibrated Confidence for large language models. Our method could serve as a strong baseline, and we hope that this work will provide some insights into the model confidence calibration. © 2024 Association for Computational Linguistics.

关键词： Calibration

来源：评论

学校读者我要写书评

暂无评论

Scale‐wise interaction fusion and knowledge distillation network for aerial scene recognition

引用

CAAI Transactions on Intelligence Technology 2023年第4期8卷 1178-1190页

作者： Hailong Ning Tao Lei Mengyuan An Hao Sun Zhanxuan Hu Asoke K.Nandi School of Computer Science and Technology Xi'an University of Posts and TelecommunicationsShaanxi Key Laboratory of Network Data Analysis and Intelligent ProcessingXi'anChina Xi'an Key Laboratory of Big Data and Intelligent Computing Xi'anChina School of Electronic Information and Artificial Intelligence Shaanxi University of Science and TechnologyXi'anChina School of Computer Central China Normal UniversityWuhanChina Department of Electronic and Electrical Engineering Brunel University LondonLondonUK Xi'an Jiaotong University Xi'anChina

Aerial scene recognition(ASR)has attracted great attention due to its increasingly essential *** of the ASR methods adopt the multi‐scale architecture because both global and local features play great roles in ***,the existing multi‐scale methods neglect the effective interactions among different scales and various spatial locations when fusing global and local features,leading to a limited ability to deal with challenges of large‐scale variation and complex background in aerial scene *** addition,existing methods may suffer from poor generalisations due to millions of to‐belearnt parameters and inconsistent predictions between global and local *** tackle these problems,this study proposes a scale‐wise interaction fusion and knowledge distillation(SIF‐KD)network for learning robust and discriminative features with scaleinvariance and background‐independent *** main highlights of this study include two *** the one hand,a global‐local features collaborative learning scheme is devised for extracting scale‐invariance features so as to tackle the large‐scale variation problem in aerial scene ***,a plug‐and‐play multi‐scale context attention fusion module is proposed for collaboratively fusing the context information between global and local *** the other hand,a scale‐wise knowledge distillation scheme is proposed to produce more consistent predictions by distilling the predictive distribution between different scales during *** experimental results show the proposed SIF‐KD network achieves the best overall accuracy with 99.68%,98.74%and 95.47%on the UCM,AID and NWPU‐RESISC45 datasets,respectively,compared with state of the arts.

关键词： deep learning image analysis image classification information fusion

来源：评论

学校读者我要写书评

暂无评论

DFA-GNN: Forward Learning of Graph Neural Networks by Direct Feedback Alignment 38

DFA-GNN: Forward Learning of Graph Neural Networks by Direct...

引用

38th Conference on Neural Information Processing Systems, NeurIPS 2024

作者： Zhao, Gongpei Wang, Tao Lang, Congyan Jin, Yi Li, Yidong Ling, Haibin Key Laboratory of Big Data & Artificial Intelligence in Transportation Ministry of Education China School of Computer Science & Technology Beijing Jiaotong University China Department of Computer Science Stony Brook University United States

Graph neural networks (GNNs) are recognized for their strong performance across various applications, with the backpropagation (BP) algorithm playing a central role in the development of most GNN models. However, despite its effectiveness, BP has limitations that challenge its biological plausibility and affect the efficiency, scalability and parallelism of training neural networks for graph-based tasks. While several non-backpropagation (non-BP) training algorithms, such as the direct feedback alignment (DFA), have been successfully applied to fully-connected and convolutional network components for handling Euclidean data, directly adapting these non-BP frameworks to manage non-Euclidean graph data in GNN models presents significant challenges. These challenges primarily arise from the violation of the independent and identically distributed (i.i.d.) assumption in graph data and the difficulty in accessing prediction errors for all samples (nodes) within the graph. To overcome these obstacles, in this paper we propose DFA-GNN, a novel forward learning framework tailored for GNNs with a case study of semi-supervised learning. The proposed method breaks the limitations of BP by using a dedicated forward training mechanism. Specifically, DFA-GNN extends the principles of DFA to adapt to graph data and unique architecture of GNNs, which incorporates the information of graph topology into the feedback links to accommodate the non-Euclidean characteristics of graph data. Additionally, for semi-supervised graph learning tasks, we developed a pseudo error generator that spreads residual errors from training data to create a pseudo error for each unlabeled node. These pseudo errors are then utilized to train GNNs using DFA. Extensive experiments on 10 public benchmarks reveal that our learning framework outperforms not only previous non-BP methods but also the standard BP methods, and it exhibits excellent robustness against various types of noise and attacks. © 2024 Neura

关键词：

来源：评论

学校读者我要写书评

暂无评论

Production and test of sPHENIX W/SciFiber electromagnetic calorimeter blocks in China

引用

Nuclear science and Techniques 2024年第8期35卷 167-176页

作者： Xiao‑Zhou Yu Xi‑Yang Wang Wei‑Hu Ma Shi‑Hong Fu Peng‑Fei Sun Jin‑Xing Song Wan‑Bing He Yang Shen Long Ma Jin‑Hui Chen Huan‑Zhong Huang Si‑Guang Wang Jing Zhou Xiao‑Mei Li Institute of Modern Physics Fudan UniversityShanghai 200433China Key Laboratory of Nuclear Physics and Ion‑Beam Application(MOE) Fudan UniversityShanghai 200433China Science and Technology on Nuclear Data Laboratory China Institute of Atomic EnergyBeijing 102413China Department of Physics and Astronomy University of CaliforniaLos AngelesCA 90095USA State Key Laboratory of Nuclear Physics and Technology School of PhysicsPeking UniversityBeijing 100871China

The sPHENIX experiment is a new generation of large acceptance detectors at the relativistic heavy ion collider at Brookhaven National laboratory,with scientific goals focusing on probing the strongly interacting Quark–Gluon plasma with hard probes of jets,open heavy flavor particles,andγ*** EMCal detector,which covers the pseudo-rapidity region of|η|≤1.1,is an essential subsystem of *** this study,we focused on producing and testing EMCal blocks covering a pseudo-rapidity of|η|∈[0.8,1.1].These,in conjunction with the central pseudo-rapidity EMCal blocks,significantly enhance the sPHENIX physics capability of the jet andγparticle *** this paper,the detector module production and testing of sPHENIX W-powder/scintillating fiber(W/ScFi)electromagnetic calorimeter blocks are *** selection of the tungsten powder,mold fabrication,QA procedures,and cosmic ray test results are discussed.

关键词： sPHENIX Tungsten/scintillating fiber electromagnetic calorimeter Cosmic ray test

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：