检索结果-内蒙古大学图书馆

Retrieve, program, repeat: Complex knowledge base question answering via alternate meta-learning

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Hua, Yuncheng Li, Yuan-Fang Haffari, Gholamreza Qi, Guilin Wu, Wei School of Computer Science and Engineering Southeast University Nanjing China Faculty of Information Technology Monash University Melbourne Australia Southeast University-Monash University Joint Research Institute Suzhou China Key Laboratory of Computer Network and Information Integration Southeast University China

A compelling approach to complex question answering is to convert the question to a sequence of actions, which can then be executed on the knowledge base to yield the answer, aka the programmer-interpreter approach. Use similar training questions to the test question, meta-learning enables the programmer to adapt to unseen questions to tackle potential distributional biases quickly. However, this comes at the cost of manually labeling similar questions to learn a retrieval model, which is tedious and expensive. In this paper, we present a novel method that automatically learns a retrieval model alternately with the programmer from weak supervision, i.e., the system’s performance with respect to the produced answers. To the best of our knowledge, this is the first attempt to train the retrieval model with the programmer jointly. Our system leads to state-of-the-art performance on a large-scale task for complex question answering over knowledge bases. We have released our code at https://***/DevinJake/MARL. Copyright © 2020, The Authors. All rights reserved.

关键词： Knowledge based systems

Cascade ResUnet with Noise Power Spectrum Loss for Low Dose CT Imaging

学校读者我要写书评

暂无评论

Cascade ResUnet with Noise Power Spectrum Loss for Low Dose ...

International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)

作者： Jin Liu Yanqin Kang Dianlin Hu Yang Chen College of Computer and Information Anhui Polytechnic University Wuhu China Key Laboratory of Computer Network and Information Integration (Southeast University) Ministry of Education Nanjing China School of Cyber Science and Engineering Southeast University Nanjing China School of Computer Science and Engineering Southeast University Nanjing China

ISBN: (数字)9780738105451

ISBN: (纸本)9781665422994

The suppression of noise-artifacts in computed tomography (CT) with a low dose scan protocol is a well-known challenge. Conventional statistical iterative algorithms are able to provide improved reconstructions, but do not work well at eliminating large streaks and strong noise. In this paper, we present a 3D cascade ResUnet neural network (Ca-ResUnet) strategy with modified noise power spectrum loss for the reduction of noise-artifacts in low dose CT reconstruction. The reconstruction workflow consists of four components: the first component is FBP (filtered backprojection) reconstruction module, the second is a ResUnet neural network with operating in the CT image, the third is image update module to compensated the loss of tiny structure, and the last is a ResUnet neural network with modified noise power spectrum loss for reconstruction image fine tuning. Verification results based on AAPM clinical data confirm that the proposed method can significantly reduce serious noise-artifacts.

关键词： Image reconstruction Computed tomography Image restoration Training Imaging X-ray imaging TV

FLSSM: A Federated Learning Storage Security Model with Homomorphic Encryption

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Li, Yang Xia, Chunhe Li, Chang Li, Xiaojian Wang, Tianbo School of Computer Science and Engineering Beihang University Beijing100191 China Key Laboratory of Beijing Network Technology Beihang University Beijing100191 China Guangxi Collaborative Innovation Center Multi-Source Information Integration and Intelligent Processing Guangxi Normal University Guilin541004 China School of Computer Science and Technology Zhengzhou University of Light Industry Zhengzhou450000 China College of Computer Science and Information Technology Guangxi Normal University Guilin541004 China School of Cyber Science and Technology Beihang University Beijing100191 China Shanghai Key Laboratory of Computer Software Evaluating and Testing Shanghai201112 China

Federated learning based on homomorphic encryption has received widespread attention due to its high security and enhanced protection of user data privacy. However, the characteristics of encrypted computation lead to three challenging problems: "computation-efficiency", "attack-tracing" and "contribution-assessment". The first refers to the efficiency of encrypted computation during model aggregation, the second refers to tracing malicious attacks in an encrypted state, and the third refers to the fairness of contribution assessment for local models after encryption. This paper proposes a federated learning storage security model with homomorphic encryption (FLSSM) to protect federated learning model privacy and address the three issues mentioned above. First, we utilize different nodes to aggregate local models in parallel, thereby improving encrypted models’ aggregation efficiency. Second, we introduce trusted supervise nodes to examine local models when the global model is attacked, enabling the tracing of malicious attacks under homomorphic encryption. Finally, we fairly reward local training nodes with encrypted local models based on trusted training time. Experiments on multiple real-world datasets show that our model significantly outperforms baseline models in terms of both efficiency and security metrics. Copyright © 2025, The Authors. All rights reserved.

关键词： Federated learning

Detecting and mitigating a sophisticated interest flooding attack in NDN from the network-wide view 1

学校读者我要写书评

暂无评论

Detecting and mitigating a sophisticated interest flooding a...

1st International Workshop on network Meets Intelligent Computations, NMIC 2019

作者： Cheng, Guang Zhao, Lixia Hu, Xiaoyan Zheng, Shaoqi Wu, Hua Li, Ruidong Fan, Chengyu School of Cyber Science and Engineering Southeast University Nanjing211189 China Key Laboratory of Computer Network and Information Integration Ministry of Education of China Southeast University China Key Laboratory of Computer Network Technology of Jiangsu Province China National Institute of Information and Communications Technology Tokyo184-8795 Japan Computer Science Department Colorado State University Fort Collins United States

ISBN: (纸本)9781728143712

Interest Flooding Attack (IFA) is one of the main security threats for the Named Data networking (NDN). Most of its existing countermeasures enable intermediate routers near the attackers to independently detect the attack and consider the typical attack scenario in which attackers directly send malicious Interests at a constant and relatively high rate. Moreover, they may also throttle legitimate Interests when enforcing the existing defence measures at intermediate routers as it is still difficult for them to distinguish the Interests issued by attackers from those issued by legitimate consumers. Instead, this work aims at a more sophisticated attack scenario in which attackers start the attack at a relatively lower rate but gradually speed up to keep the Pending Interest Tables (PITs) of the victims increasing to finally deplete the PIT resources for legitimate consumers. It is relatively difficult for intermediate routers to independently and timely detect such a sophisticated IFA. To solve this problem, we propose a mechanism to detect the sophisticated IFA from the network-wide view. A central controller monitors the network and makes a comprehensive and prompt decision on whether there is an ongoing IFA based on the overall state of the whole network collected from the abnormity information reports sent by the first-hop routers of attackers. Attack sources can be directly located after an IFA is determined and then the malicious Interests can be prevented from entering the network without throttling legitimate Interests. We conduct an experimental study to evaluate the performance of the proposed mechanism and explore the parameter settings of the attack detection algorithm at access routers. The experimental results validate that our mechanism can timely detect and mitigate the sophisticated IFA without throttling requests from legitimate consumers. © 2019 IEEE.

关键词： computer network security Internet telecommunication network routing Attackers network security telecommunication network routing NDN gene incomplete Freund's adjuvant Interests Internets SYN flood THROTTLE VALVE Routers parameter setting

dptCry: an approach to decrypting ransomware WannaCry based on API hooking

学校读者我要写书评

暂无评论

CCF Transactions on networking 2019年第3-4期2卷 207-216页

作者： Guang Cheng Chunsheng Guo Yongning Tang Southeast University Key Lab of Computer Network and Information Integration (Southeast University) Ministry of Education Illinois State University

Neural topic modeling with bidirectional adversarial training

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Wang, Rui Hu, Xuemeng Zhou, Deyu He, Yulan Xiong, Yuxuan Ye, Chenchen Xu, Haiyang School of Computer Science and Engineering Key Laboratory of Computer Network and Information Integration Ministry of Education Southeast University China Department of Computer Science University of Warwick United Kingdom AI Labs Didi Chuxing Co. Ltd. Beijing China

Recent years have witnessed a surge of interests of using neural topic models for automatic topic extraction from text, since they avoid the complicated mathematical derivations for model inference as in traditional topic models such as Latent Dirichlet Allocation (LDA). However, these models either typically assume improper prior (e.g. Gaussian or Logistic Normal) over latent topic space or could not infer topic distribution for a given document. To address these limitations, we propose a neural topic modeling approach, called Bidirectional Adversarial Topic (BAT) model, which represents the first attempt of applying bidirectional adversarial training for neural topic modeling. The proposed BAT builds a two-way projection between the document-topic distribution and the document-word distribution. It uses a generator to capture the semantic patterns from texts and an encoder for topic inference. Furthermore, to incorporate word relatedness information, the Bidirectional Adversarial Topic model with Gaussian (Gaussian-BAT) is extended from BAT. To verify the effectiveness of BAT and Gaussian-BAT, three benchmark corpora are used in our experiments. The experimental results show that BAT and Gaussian-BAT obtain more coherent topics, outperforming several competitive baselines. Moreover, when performing text clustering based on the extracted topics, our models outperform all the baselines, with more significant improvements achieved by Gaussian-BAT where an increase of near 6% is observed in accuracy. Copyright © 2020, The Authors. All rights reserved.

关键词： Gaussian distribution

Learning Expectation of Label Distribution for Facial Age and Attractiveness Estimation

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Gao, Bin-Bin Liu, Xin-Xin Zhou, Hong-Yu Wu, Jianxin Geng, Xin YouTu Lab Tencent Shenzhen518075 China National Key Laboratory for Novel Software Technology Nanjing University Nanjing210023 China MOE Key Laboratory of Computer Network and Information Integration Southeast University Nanjing211189 China

Facial attributes (e.g., age and attractiveness) estimation performance has been greatly improved by using convolutional neural networks. However, existing methods have an inconsistency between the training objectives and the evaluation metric, so they may be suboptimal. In addition, these methods always adopt image classification or face recognition models with a large amount of parameters, which carry expensive computation cost and storage overhead. In this paper, we firstly analyze the essential relationship between two state-of-the-art methods (Ranking-CNN and DLDL) and show that the Ranking method is in fact learning label distribution implicitly. This result thus firstly unifies two existing popular state-of-the-art methods into the DLDL framework. Second, in order to alleviate the inconsistency and reduce resource consumption, we design a lightweight network architecture and propose a unified framework which can jointly learn facial attribute distribution and regress attribute value. The effectiveness of our approach has been demonstrated on both facial age and attractiveness estimation tasks. Our method achieves new state-of-the-art results using the single model with 36× fewer parameters and 3× faster inference speed on facial age/attractiveness estimation. Moreover, our method can achieve comparable results as the state-of-the-art even though the number of parameters is further reduced to 0.9M (3.8MB disk storage). Copyright © 2020, The Authors. All rights reserved.

关键词： Convolutional neural networks

A Topology-Adaptive Deep Model for Power System Multifault Diagnosis

学校读者我要写书评

暂无评论

A Topology-Adaptive Deep Model for Power System Multifault D...

IEEE International Conference on Systems, Man and Cybernetics

作者： Yangyang Liu Aibo Song Jiahui Jin Mingyu zhai Yingying Xue Feng Li School of Computer Science and Engineering Southeast University Nanjing China Key Laboratory of Computer Network and Information Integration Southeast University Ministry of Education China NARI Group Corporation/State Grid Electric Power Research Institute Nanjing China

ISBN: (数字)9781728185262

ISBN: (纸本)9781728185279

Quickly identifying faulty sections is tremendously important for power systems, yet challenging due to handling the variations of complex alarm patterns. Existing works have focused on finding fault section clues solely from alarm information (and ignoring power system topology information). So they are only sensitive to alarms from power systems with pre-assumed topology structures, and encounter difficulties when a system's topology changes. To adapt to unknown or varying system topologies, here we present a Topology-Adaptive Deep Model (TADM) for power system multifault diagnosis. TADM mines the underlying mapping from alarm and topology information to each section's fault status. It consists of a deep iterative network (DIN), a one-layer fully connected network (FCN), and section-wise multifault diagnosis (SWMD) subnetwork. TADM first models a fault power system as a graph, from which DIN iteratively integrates the alarm and topology information in the region from each node to its T -hop neighbors, and learns their local correlation. Limited to T 's size, FCN then combines all local correlations to determine the global correlation between alarm and topology information across the entire power system. To implement multifault diagnosis, learned local and global correlations serve as topology-related fault representations for input as an SWMD (to predict all sections' fault states one by one). A comprehensive experimental study demonstrates that TADM outperforms state-of-the-art models in both multifault diagnosis and adapting to system topologies. The source code of the TADM is available onlline 1 .

关键词： Knowledge engineering Adaptation models Correlation network topology Social networking (online) Topology Power systems

Less is More: Data-Efficient Complex Question Answering over Knowledge Bases

学校读者我要写书评

暂无评论

arXiv 2020年

作者： Hua, Yuncheng Li, Yuan-Fang Qi, Guilin Wu, Wei Zhang, Jingyao Qi, Daiqing School of Computer Science and Engineering Southeast University Nanjing China Faculty of Information Technology Monash University Melbourne Australia Key Laboratory of Computer Network and Information Integration Southeast University Ministry of Education Nanjing China Southeast University Monash University Joint Research Institute Suzhou China

Question answering is an effective method for obtaining information from knowledge bases (KB). In this paper, we propose the Neural-Symbolic Complex Question Answering (NS-CQA) model, a data-efficient reinforcement learning framework for complex question answering by using only a modest number of training samples. Our framework consists of a neural generator and a symbolic executor that, respectively, transforms a natural-language question into a sequence of primitive actions, and executes them over the knowledge base to compute the answer. We carefully formulate a set of primitive symbolic actions that allows us to not only simplify our neural network design but also accelerate model convergence. To reduce search space, we employ the copy and masking mechanisms in our encoder-decoder architecture to drastically reduce the decoder output vocabulary and improve model generalizability. We equip our model with a memory buffer that stores high-reward promising programs. Besides, we propose an adaptive reward function. By comparing the generated trial with the trials stored in the memory buffer, we derive the curriculum-guided reward bonus, i.e., the proximity and the novelty. To mitigate the sparse reward problem, we combine the adaptive reward and the reward bonus, reshaping the sparse reward into dense feedback. Also, we encourage the model to generate new trials to avoid imitating the spurious trials while making the model remember the past high-reward trials to improve data efficiency. Our NS-CQA model is evaluated on two datasets: CQA, a recent large-scale complex question answering dataset, and WebQuestionsSP, a multi-hop question answering dataset. On both datasets, our model outperforms the state-of-the-art models. Notably, on CQA, NS-CQA performs well on questions with higher complexity, while only using approximately 1% of the total training samples. Copyright © 2020, The Authors. All rights reserved.

关键词： Reinforcement learning