检索结果-内蒙古大学图书馆

Discriminative features matter: Multi-layer bilinear pooling for camera localization 30

学校读者我要写书评

暂无评论

Discriminative features matter: Multi-layer bilinear pooling...

30th British Machine Vision Conference, BMVC 2019

作者： Wang, Xin Wang, Xiang Wang, Chen Bai, Xiao Wu, Jing Hancock, Edwin Robert School of Computer Science and Engineering Beijing Advanced Innovation Center for Big Data and Brain Computing Jiangxi Research Institute Beihang University Beijing China School of Computer Science and Informatics Cardiff University Cardiff United Kingdom Department of Computer Science University of York York United Kingdom

Deep learning based camera localization from a single image has been explored recently since these methods are computationally efficient. However, existing methods only provide general global representations, from which an accurate pose estimation can not be reliably derived. We claim that effective feature representations for accurate pose estimation shall be both "informative" (focusing on geometrically meaningful regions) and "discriminative" (accounting for different poses of similar images). Therefore, we propose a novel multi-layer factorized bilinear pooling module for feature aggregation. Specifically, informative features are selected via bilinear pooling, and discriminative features are highlighted via multi-layer fusion. We develop a new network for camera localization using the proposed feature pooling module. The effectiveness of our approach is demonstrated by experiments on an outdoor Cambridge Landmarks dataset and an indoor 7 Scenes dataset. The results show that focusing on discriminative features significantly improves the network performance of camera localization in most cases. Codes will be available soon. © 2019. The copyright of this document resides with its authors.

关键词： Deep learning

Efficient Bitcoin Address Classification Using Quantum-Inspired Feature Selection

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Sie, Ming-Fong Chang, Yen-Jui Lin, Chien-Lung Chang, Ching-Ray Liao, Shih-Wei Department of Computer Science and Information Engineering National Taiwan University No. 1 Sec. 4 Roosevelt Rd. Taipei City10617 Taiwan Master Program in Intelligent Computing and Big Data Chung Yuan Christian University No. 499 Xingzhong N. Rd. Taoyuan City320317 Taiwan Department of Physics National Taiwan University Taiwan Quantum Information Center Chung Yuan Christian University No. 499 Xingzhong N. Rd. Taoyuan City320317 Taiwan

Over 900 million Bitcoin transactions have been recorded, posing considerable challenges for machine learning regarding computation time and maintaining prediction accuracy. We propose an innovative approach using quantum-inspired algorithms implemented with Simulated Annealing and Quantum Annealing to address the challenge of local minima in solution spaces. This method efficiently identifies key features linked to mixer addresses, significantly reducing model training time. By categorizing Bitcoin addresses into six classes—exchanges, faucets, gambling, marketplaces, mixers, and mining pools—and applying supervised learning methods, our results demonstrate that feature selection with SA reduced training time by 30.3% compared to using all features in a random forest model while maintaining a 91% F1-score for mixer addresses. This highlights the potential of quantum-inspired algorithms to swiftly and accurately identify high-risk Bitcoin addresses based on transaction features. © 2024, CC BY.

关键词： Simulated annealing

P-FedAvg: Parallelizing federated learning with theoretical guarantees 40

学校读者我要写书评

暂无评论

P-FedAvg: Parallelizing federated learning with theoretical ...

40th IEEE Conference on computer Communications, INFOCOM 2021

作者： Zhong, Zhicong Zhou, Yipeng Wu, Di Chen, Xu Chen, Min Li, Chao Sheng, Quan Z. Sun Yat-Sen University School of Computer Science and Engineering Guangzhou510006 China Guangdong Key Laboratory of Big Data Analysis and Processing Guangzhou510006 China PCL Research Center of Networks and Communications Peng Cheng Laboratory Shenzhen518000 China Department of Computing Macquarie University SydneyNSW2109 Australia School of Computer Science and Technology Huazhong University of Science and Technology Wuhan430074 China Tencent Technology Co. Ltd Shenzhen China

ISBN: (纸本)9780738112817

With the growth of participating clients, the centralized parameter server (PS) will seriously limit the scale and efficiency of Federated Learning (FL). A straightforward approach to scale up the FL system is to construct a Parallel FL (PFL) system with multiple PSes. However, it is unclear whether PFL can really achieve a faster convergence rate or not. Even if the answer is yes, it is non-trivial to design a highly efficient parameter average algorithm for a PFL system. In this paper, we propose a completely parallelizable FL algorithm called P-FedAvg under the PFL architecture. P-FedAvg extends the well-known FedAvg algorithm by allowing multiple PSes to cooperate and train a learning model together. In P-FedAvg, each PS is only responsible for a fraction of total clients, but PSes can mix model parameters in a dedicatedly designed way so that the FL model can well converge. Different from heuristic-based algorithms, P-FedAvg is with theoretical guarantees. To be rigorous, we conduct theoretical analysis on the convergence rate of P-FedAvg, and derive the optimal weights for each PS to mix parameters with its neighbors. We also examine how the overlay topology formed by PSes affects the convergence rate and robustness of a PFL system. Lastly, we perform extensive experiments with real datasets to verify our analysis and demonstrate that P-FedAvg can significantly improve convergence rates than traditional FedAvg and other competitive baselines. We believe that our work can help to lay a theoretical foundation for building more efficient PFL systems. © 2021 IEEE.

关键词： Network topology

ImplantFormer: Vision Transformer based Implant Position Regression Using Dental CBCT data

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Yang, Xinquan Li, Xuguang Li, Xuechen Wu, Peixi Shen, Linlin Deng, Yongqiang College of Computer Science and Software Engineering Shenzhen University Shenzhen China AI Research Center for Medical Image Analysis and Diagnosis Shenzhen University Shenzhen China National Engineering Laboratory for Big Data System Computing Technology Shenzhen University China Department of Stomatology Shenzhen University General Hospital Shenzhen China The School of Dentistry Shenzhen University Shenzhen China

Implant prosthesis is the most appropriate treatment for dentition defect or dentition loss, which usually involves a surgical guide design process to decide the implant position. However, such design heavily relies on the subjective experiences of dentists. In this paper, a transformer-based Implant Position Regression Network, ImplantFormer, is proposed to automatically predict the implant position based on the oral CBCT data. We creatively propose to predict the implant position using the 2D axial view of the tooth crown area and fit a centerline of the implant to obtain the actual implant position at the tooth root. Convolutional stem and decoder are designed to coarsely extract image features before the operation of patch embedding and integrate multi-level feature maps for robust prediction, respectively. As both long-range relationship and local features are involved, our approach can better represent global information and achieves better location performance. Extensive experiments on a dental implant dataset through five-fold cross-validation demonstrated that the proposed ImplantFormer achieves superior performance than existing methods. Copyright © 2022, The Authors. All rights reserved.

关键词： Forecasting

An Accurate and Explainable Deep Learning System Improves Interobserver Agreement in the Interpretation of Chest Radiograph

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Pham, Huy Hieu Nguyen, Ha Q. Nguyen, Hieu T. Le, Linh T. Khanh, Lam College of Engineering and Computer Science VinUniversity Hanoi Viet Nam VinUni-Illinois Smart Health Center VinUniversity Hanoi Viet Nam Smart Health Center VinBigData JSC Hanoi Viet Nam Vingroup Big Data Institute Hanoi Viet Nam Hanoi Medical University Hospital Department of Radiology Hanoi Viet Nam Hospital 108 Department of Radiology Hanoi Viet Nam

Interpretation of chest radiographs (CXR) is a difficult but essential task for detecting thoracic abnormalities. Recent artificial intelligence (AI) algorithms have achieved radiologist-level performance on various medical classification tasks. However, only a few studies addressed the localization of abnormal findings from CXR scans, which is essential in explaining the image-level classification to radiologists. Additionally, the actual impact of AI algorithms on the diagnostic performance of radiologists in clinical practice remains relatively unclear. To bridge these gaps, we developed an explainable deep learning system called VinDr-CXR that can classify a CXR scan into multiple thoracic diseases and, at the same time, localize most types of critical findings on the image. VinDr-CXR was trained on 51,485 CXR scans with radiologist-provided bounding box annotations. It demonstrated a comparable performance to experienced radiologists in classifying 6 common thoracic diseases on a retrospective validation set of 3,000 CXR scans, with a mean area under the receiver operating characteristic curve (AUROC) of 0.967 (95% confidence interval [CI]: 0.958–0.975). The VinDr-CXR was also externally validated in independent patient cohorts and showed its robustness. For the localization task with 14 types of lesions, our free-response receiver operating characteristic (FROC) analysis showed that the VinDr-CXR achieved a sensitivity of 80.2% at the rate of 1.0 false-positive lesion identified per scan. A prospective study was also conducted to measure the clinical impact of the VinDr-CXR in assisting six experienced radiologists. The results indicated that the proposed system, when used as a diagnosis supporting tool, significantly improved the agreement between radiologists themselves with an increase of 1.5% in mean Fleiss’ Kappa. We also observed that, after the radiologists consulted VinDr-CXR’s suggestions, the agreement between each of them and the system was remarkab

关键词： Object detection

Neuromodulated Meta-Learning

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Wang, Jingyao Guo, Huijie Qiang, Wenwen Li, Jiangmeng Zheng, Changwen Xiong, Hui Hua, Gang University of Chinese Academy of Sciences Beijing China National Key Laboratory of Space Integrated Information System Institute of Software Chinese Academy of Sciences Beijing China Beijing Advanced Innovation Center for Big Data and Brain Computing Beihang University Beijing China Thrust of Artificial Intelligence Hong Kong University of Science and Technology Guangzhou China Department of Computer Science and Engineering The Hong Kong University of Science and Technology Hong Kong Wormpex AI Research LLC BellevueWA98004 United States

Humans excel at adapting perceptions and actions to diverse environments, enabling efficient interaction with the external world. This adaptive capability relies on the biological nervous system (BNS), which activates different brain regions for distinct tasks. Meta-learning similarly trains machines to handle multiple tasks but relies on a fixed network structure, not as flexible as BNS. To investigate the role of flexible network structure (FNS) in meta-learning, we conduct extensive empirical and theoretical analyses, finding that model performance is tied to structure, with no universally optimal pattern across tasks. This reveals the crucial role of FNS in meta-learning, ensuring meta-learning to generate the optimal structure for each task, thereby maximizing the performance and learning efficiency of meta-learning. Motivated by this insight, we propose to define, measure, and model FNS in meta-learning. First, we define that an effective FNS should possess frugality, plasticity, and sensitivity. Then, to quantify FNS in practice, we present three measurements for these properties, collectively forming the structure constraint with theoretical supports. Building on this, we finally propose Neuromodulated Meta-Learning (NeuronML) to model FNS in meta-learning. It utilizes bi-level optimization to update both weights and structure with the structure constraint. Extensive theoretical and empirical evaluations demonstrate the effectiveness of NeuronML on various tasks. Code is publicly available at https://***/WangJingyao07/NeuronML. Copyright © 2024, The Authors. All rights reserved.

关键词： Contrastive Learning

Explainable machine learning models for estimating daily dissolved oxygen concentration of the Tualatin River

学校读者我要写书评

暂无评论

engineering Applications of Computational Fluid Mechanics 2024年第1期18卷

作者： Li, Shuguang Qasem, Sultan Noman Band, Shahab S. Ameri, Rasoul Pai, Hao-Ting Mehdizadeh, Saeid School of Computer Science and Technology Shandong Technology and Business University Yantai China Computer Science Department College of Computer and Information Sciences Imam Mohammad Ibn Saud Islamic University (IMSIU) Riyadh Saudi Arabia Computer Science Department Faculty of Applied Science Taiz University Taiz Yemen Future Technology Research Center National Yunlin University of Science and Technology Douliu Taiwan Department of Information Management International Graduate School of Artificial Intelligence National Yunlin University of Science and Technology Douliu Taiwan Department of Information Management National Yunlin University of Science and Technology Douliu Taiwan Bachelor Program of Big Data Applications in Business National Pingtung University Pingtung Taiwan Water Engineering Department Urmia University Urmia Iran

Monitoring the quality of river water is of fundamental importance and needs to be taken into consideration when it comes to the research into the hydrological field. In this context, the concentration of the dissolved oxygen (DO) is one of the most significant indicators of the quality of river water. The current study aimed to estimate the minimum, maximum, and mean DO concentrations (DO min, DO max, DO mean) at a gauging station located on Tualatin River, United States. To that end, four machine learning models, such as support vector regression (SVR), multi-layer perceptron (MLP), random forest (RF), and gradient boosting (GB) were established. Root mean square error (RMSE), mean absolute error (MAE), coefficient of correlation (R), and Nash-Sutcliffe efficiency (NSE) metrics were employed to better assess the accuracies of these models. The modeling results demonstrated that the SVR and MLP surpassed the RF and GB models. Despite this, the SVR was concluded to be the best-performing method when used to estimate the DO min, DO max, and DO mean. The best error statistics in the testing phase were related to the SVR model with full (four) inputs to estimate DO mean concentration (RMSE = 0.663 mg/l, MAE = 0.508 mg/l, R = 0.945, NSE = 0.875). Finally, the explainability of the superior models (i.e. SVR models) was conducted using SHapley Additive exPlanations (SHAP) for the first time to estimate DO concentration. In fact, evaluating the explainability of machine learning models can provide useful information about the impact of each of the input estimators used in the procedure of models development. It was concluded that the specific conductance (SC) and followed by water temperature (WT) could provide the most contributions for estimating the DO min, DO max, and DO mean concentrations. © 2024 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group.

关键词： dissolved oxygen concentration estimation Explainable machine learning SHapley additive explanations

Near-optimal representation learning for linear bandits and linear RL

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Hu, Jiachen Chen, Xiaoyu Jin, Chi Li, Lihong Wang, Liwei Key Laboratory of Machine Perception MOE School of EECS Peking University China Department of Electrical and Computer Engineering Princeton University United States Amazon United States Center for Data Science Peking University Beijing Institute of Big Data Research China

This paper studies representation learning for multi-task linear bandits and multi-task episodic RL with linear value function approximation. We first consider the setting where we play M linear bandits with dimension d concurrently, and these bandits share a common k-dimensional linear representation so that k d and k M. We propose a sample-efficient algorithm, MTLR-OFUL, which leverages the shared representation to achieve Õ(M√dkT + d√kMT) regret, with T being the number of total steps. Our regret significantly improves upon the baseline Õ(Md√T) achieved by solving each task independently. We further develop a lower bound that shows our regret is near-optimal when d > M. Furthermore, we extend the algorithm and analysis to multi-task episodic RL with linear value function approximation under low inherent Bellman error (Zanette et al., 2020a). To the best of our knowledge, this is the first theoretical result that characterizes the benefits of multi-task representation learning for exploration in RL with function approximation. Copyright © 2021, The Authors. All rights reserved.

关键词： Machine learning

A Text Multi-label Classification Scheme Based on Resampling and Ensemble Learning 8th

学校读者我要写书评

暂无评论

A Text Multi-label Classification Scheme Based on Resampling...

8th International Conference on Artificial Intelligence and Security , ICAIS 2022

作者： Wang, Tianhao Weng, Tianrang Ji, Jiacheng Zhong, Mingjun Zhang, Baili School of Computer Science and Engineering Southeast University Nanjing211189 China Key Laboratory of Computer Network and Information Integration in Southeast University Ministry of Education Nanjing211189 China Research Center for Judicial Big Data Supreme Count of China Nanjing211189 China Department of Computing Science University of Aberdeen AberdeenAB24 3UE United Kingdom

ISBN: (纸本)9783031067600

The medical dispute cases are professional and closely related to medicine. Therefore, the mediation of cases in practice depends heavily on similar historical cases. Multi-label classification of legal documents can efficiently filter irrelevant historical cases, which helps to recommend the similar historical cases faster and better. However, the imbalance and label symbiosis of the data set directly affect multi-label classification of legal documents. Therefore, a multi-label classification scheme based on resampling and ensemble learning is presented in this paper The scheme includes two parts: in the first part, in order to reduce the impact of label symbiosis on resampling, a resampling algorithm based on the average sparsity of the label set is proposed improve the imbalance of the data set;in the second one, a multi-label classification algorithm based on ensemble learning is proposed to train multiple base classifiers and combine each base classifier with a voting strategy of one vote. It can effectively improve the effect of multi-label classification. The experimental results show that the scheme proposed in this paper can improve the effect of multi-label classification and is not only suitable for legal documents but also applicable for other text data sets with imbalanced classes and label symbiosis problems. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Learning algorithms