检索结果-内蒙古大学图书馆

41st International Conference on Machine Learning, ICML 2024

作者： Ji, Tianying Luo, Yu Sun, Fuchun Zhan, Xianyuan Zhang, Jianwei Xu, Huazhe Department of Computer Science and Technology Tsinghua University China Institute for AI Industry Research Tsinghua University China Shanghai Artificial Intelligence Laboratory China Department of Informatics University of Hamburg Germany Institute for Interdisciplinary Information Sciences Tsinghua University China Shanghai Qi Zhi Institute China

Learning high-quality Q-value functions plays a key role in the success of many modern off-policy deep reinforcement learning (RL) algorithms. Previous works primarily focus on addressing the value overestimation issue, an outcome of adopting function approximators and off-policy learning. Deviating from the common viewpoint, we observe that Q-values are often underestimated in the latter stage of the RL training process, potentially hindering policy learning and reducing sample efficiency. We find that such a long-neglected phenomenon is often related to the use of inferior actions from the current policy in Bellman updates as compared to the more optimal action samples in the replay buffer. To address this issue, our insight is to incorporate sufficient exploitation of past successes while maintaining exploration optimism. We propose the Blended Exploitation and Exploration (BEE) operator, a simple yet effective approach that updates Qvalue using both historical best-performing actions and the current policy. Based on BEE, the resulting practical algorithm BAC outperforms state-of-the-art methods in over 50 continuous control tasks and achieves strong performance in failure-prone scenarios and real-world robot tasks. Benchmark results and videos are available at https://***/BEE/. Copyright 2024 by the author(s)

关键词： Deep reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Offline-Boosted Actor-Critic: Adaptively Blending Optimal Historical Behaviors in Deep Off-Policy RL 41

Offline-Boosted Actor-Critic: Adaptively Blending Optimal Hi...

引用

41st International Conference on Machine Learning, ICML 2024

作者： Luo, Yu Ji, Tianying Sun, Fuchun Zhang, Jianwei Xu, Huazhe Zhan, Xianyuan Department of Computer Science and Technology Tsinghua University China Department of Informatics University of Hamburg Germany Institute for Interdisciplinary Information Sciences Tsinghua University China Shanghai Qi Zhi Institute China Shanghai Artificial Intelligence Laboratory China Institute for AI Industry Research Tsinghua University China

Off-policy reinforcement learning (RL) has achieved notable success in tackling many complex real-world tasks, by leveraging previously collected data for policy learning. However, most existing off-policy RL algorithms fail to maximally exploit the information in the replay buffer, limiting sample efficiency and policy performance. In this work, we discover that concurrently training an offline RL policy based on the shared online replay buffer can sometimes outperform the original online learning policy, though the occurrence of such performance gains remains uncertain. This motivates a new possibility of harnessing the emergent outperforming offline optimal policy to improve online policy learning. Based on this insight, we present Offline-Boosted Actor-Critic (OBAC), a model-free online RL framework that elegantly identifies the outperforming offline policy through value comparison, and uses it as an adaptive constraint to guarantee stronger policy learning performance. Our experiments demonstrate that OBAC outperforms other popular model-free RL baselines and rivals advanced model-based RL methods in terms of sample efficiency and asymptotic performance across 53 tasks spanning 6 task suites1 Copyright 2024 by the author(s)

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Text Analysis and Recognition of Emotional Content Using Deep Learning Methods and BERT

Text Analysis and Recognition of Emotional Content Using Dee...

引用

International Conference on computer and Information Science (ACIS)

作者： Efthymios Andrikakis Isidoros Perikos Michael Paraskevas Ioannis Hatzilygeroudis Department of Computer Engineering & Informatics University of Patras Patras Greece Department of Computer Engineering & Informatics University of Patras and Computer Technology Institute and Press “Diophantus” Patras Greece

Recognizing the emotional content of Natural Language sentences can improve the way humans communicate with a computer system by enabling them to recognize and imitate emotional expressions. In this paper, deep learning, deep neural networks as well as Transformers were examined. Specifically, we designed and developed deep learning methods and BERT-based implementations for recognizing emotional content in user-generated data. Extensive experiments were conducted using these models on a variety of textual data and all the designed methods were evaluated. The results of the study show that BERT achieved the best overall performance, getting an F1-score of 0.86 on the Twitter Data dataset. Also, the Bi-LSTM with Attention Mechanism and Bi-LSTM using Word2vec performed quite well, achieving F1-scores of 0.83 and 0.82, respectively.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Ant Colony Optimization for Service Composition in Multi-cloud Environments 15

Ant Colony Optimization for Service Composition in Multi-clo...

引用

15th International Conference on Computing Communication and Networking Technologies, ICCCNT 2024

作者： Maurya, Mahesh Vidhya, G. Akram, Faiz Saravanan, M. Venkata Ramana, K. Rajakumari, S. K.C. College of Engineering Department of Computer Engineering Mith Bunder road Thane East Maharashtra Kopri India Sona College of Technology Department of Computer Science and Engineering Tamilnadu Salem India Jimma Institute of Technology Jimma University Faculty of Computing and Informatics Department of Computer Science Oromia Region Jimma Ethiopia Sri Eshwar College of Engineering Department of Ece Tamilnadu Coimbatore India Qis College of Engineering and Technology Department of Cse Andrapradesh Ongole India R.M.D. Engineering College Department of Mathematics India

ISBN: (纸本)9798350370249

the Ant Colony Optimization (ACO) algorithm is a famous metaheuristic technique that has been successfully applied in various optimization issues. It mimics the foraging conduct of ants to locate the shortest path among meals assets and their colony. Recently, ACO has been increasingly more applied for provider composition in multi-cloud environments. This involves the composition of offerings from different cloud carriers to satisfy the service necessities of a person. By way of using ACO, services can be correctly selected and composed from a couple of carriers, considering elements including cost, excellent, and availability. ACO has proven promising outcomes in enhancing the overall performance and reliability of carrier composition in multi-cloud environments, making it a precious technique in todays tremendously distributed computing landscape. © 2024 IEEE.

关键词： Distributed cloud

来源：评论

学校读者我要写书评

暂无评论

OMPO: A Unified Framework for RL under Policy and Dynamics Shifts 41

OMPO: A Unified Framework for RL under Policy and Dynamics S...

引用

41st International Conference on Machine Learning, ICML 2024

Training reinforcement learning policies using environment interaction data collected from varying policies or dynamics presents a fundamental challenge. Existing works often overlook the distribution discrepancies induced by policy or dynamics shifts, or rely on specialized algorithms with task priors, thus often resulting in suboptimal policy performances and high learning variances. In this paper, we identify a unified strategy for online RL policy learning under diverse settings of policy and dynamics shifts: transition occupancy matching. In light of this, we introduce a surrogate policy learning objective by considering the transition occupancy discrepancies and then cast it into a tractable min-max optimization problem through dual reformulation. Our method, dubbed Occupancy-Matching Policy Optimization (OMPO), features a specialized actor-critic structure equipped with a distribution discriminator and a small-size local buffer. We conduct extensive experiments based on the OpenAI Gym, Meta-World, and Panda Robots environments, encompassing policy shifts under stationary and non-stationary dynamics, as well as domain adaption. The results demonstrate that OMPO outperforms the specialized baselines from different categories in all settings. We also find that OMPO exhibits particularly strong performance when combined with domain randomization, highlighting its potential in RL-based robotics applications1 Copyright 2024 by the author(s)

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Classifying Upper Extremity Motor Function Using Deep Learning and Third-Person Video Data

Classifying Upper Extremity Motor Function Using Deep Learni...

引用

2024 IEEE International Conference on Big Data, BigData 2024

作者： Tran, Tan Chang, Lin-Ching Lum, Peter The Catholic University of America Department of Electrical Engineering and Computer Science WashingtonDC United States New Jersey Institute of Technology Department of Informatics NewarkNJ United States The Catholic University of America Department of Biomedical Engineering WashingtonDC United States

ISBN: (纸本)9798350362480

Post-stroke rehabilitation of upper extremity (UE) motor function is essential. Despite the widespread use of UE rehabilitation in clinical settings, assessing the success of these treatments is challenging. Methods for evaluating UE motor function include task performance under clinical supervision, patient self-reports, and data analysis from wearable devices equipped with accelerometers and gyroscopes. In prior research, we demonstrated that machine learning and deep learning models using data from a single wrist-worn accelerometer sensor could accurately differentiate between functional and non-functional UE movements in stroke patients. To overcome the limitations of single wrist-worn accelerometer sensors - challenges in capturing the full context of functional movements - this study presents a new deep learning (DL) based framework designed to classify functional and non-functional arm movements in videos captured from a conventional camera. The system is entirely automated and comprises two DL networks. The first network performs human pose estimation, extracting 2D pose key points of the paretic arm(s) and torso on 2-second sequences of the frames. The second network then uses these 2D pose landmarks to classify the movement sequence as functional or non-functional. This system offers two key benefits for rehabilitation. First, it can automatically generate initial annotations for video frames, significantly reducing the time needed for manual labeling. Second, analyzing the effectiveness of UE rehabilitation through third-person videos allows for objective outcome measurement of UE treatments in stroke survivors' home environments. © 2024 IEEE.

关键词： Accelerometers

来源：评论

学校读者我要写书评

暂无评论

A Comparative Analysis of Tweet Analysis Algorithms Using Natural Language Processing and Machine Learning Models 18

A Comparative Analysis of Tweet Analysis Algorithms Using Na...

引用

18th International Workshop on Semantic and Social Media Adaptation and Personalization, SMAP 2023

作者： Kanavos, Andreas Antonopoulos, Nikos Karamitsos, Ioannis Mylonas, Phivos Ionian University Department of Informatics Corfu Greece Ionian University Department of Digital Media and Communication NeMeCULAB Argostoli Greece Rochester Institute of Technology Department of Graduate and Research Dubai United Arab Emirates University of West Attica Department of Informatics and Computer Engineering Athens Greece

ISBN: (纸本)9798350327717

Online Social Networks (OSNs) have become integral platforms for information sharing, attracting both legitimate users and spammers. Detecting and mitigating spam within these networks pose significant challenges due to the dynamic nature of content and user behavior. In this paper, we present a comprehensive comparative analysis of algorithms for tweet analysis, focusing on Natural Language Processing (NLP) and Machine Learning (ML) models. We evaluate these algorithms through sentiment analysis and multiple attribute analysis, utilizing diverse methodologies and datasets. Our study explores feed-forward neural networks, Bayesian classifiers, and transformer-based models for NLP tasks, achieving high prediction accuracy and insightful metrics such as precision, recall, and F1 score. Furthermore, we delve into multiple attribute analysis using Random Forest, Logistic Regression, and Gradient Boosting algorithms. Through a systematic exploration of various approaches, this work contributes to a deeper understanding of spam detection and sentiment analysis within the context of OSNs, paving the way for enhanced social network security and content analysis. © 2023 IEEE.

关键词： Social networking (online)

来源：评论

学校读者我要写书评

暂无评论

Tangles in IOTA to make crypto currency transactions free and secure

引用

International Journal of Cloud Computing 2022年第1期11卷 1-13页

作者： Anakath, A.S. Kannadasan, R. Prabakaran, N. Basha, M.S. Saleem Department of MCA EGS Pillai Engineering College Nagapattinam 611002 India School Computer Science and Engineering Vellore Institute of Technology Vellore632014 India Department of Computing and Informatics Mazoon College Muscat133 Oman

Blockchain introduction has made a revolutionary change in the crypto currency around the world but it has not delivered on its promises of free and faster transaction confirmation. Serguei Popov's proposal of using tangles, a directed acyclic graph, which essentially is considered to be the successor of blockchains and offers the required features like machine to machine micro payment and fee-less transaction. It requires the user to approve the previous two transactions in the web to participate in the network. This essentially eliminates the miners and the mining part form the currency exchange and provides the user or participants to do their transactions fee-less. Since the participant verifies the previous two transactions it also contributes to the security of the tangle. In this paper, features of internet of things application (IOTA) and all the improvements in it using tangles are discussed along with how it contributed to the security and how it enables the participants to have fee-less transactions is also discussed. Copyright © 2022 Inderscience Enterprises Ltd.

关键词： Distributed ledger

来源：评论

学校读者我要写书评

暂无评论

SAR Images Change Detection Based on Attention Mechanism-Convolutional Wavelet Neural Network

引用

IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 2025年 18卷 12133-12149页

作者： Jiahui, J.E. Wang, Lu Zhao, Chunhui Mathiopoulos, P. Takis Ohtsuki, Tomoaki Adachi, Fumiyuki Harbin Engineering University College of Information and Communication Engineering Harbin150001 China Ministry of Industry and Information Technology Key Laboratory of Advanced Marine Communication and Information Technology Harbin150001 China National and Kapodistrian University of Athens Department of Informatics and Telecommunications 15772 Greece Keio University Department of Information and Computer Science Kanagawa 2238522 Japan Tohoku University International Research Institute of Disaster Science Sendai9808576 Japan

As synthetic aperture radar (SAR) image change detection can continuously acquire target information under all weather conditions, it has been used in the past for various applications. However, it is very challenging to use SAR images for small area change detection due to the presence of noise and the imbalance between classes. To deal with these problems this article proposes a SAR images change detection scheme which is based upon an Attention Mechanism and Convolutional Wavelet Neural Network. First, employing Multiscale Superpixel Reconstructed Difference Image effectively enhances the edge information of the images. Second, a two-stage Weighted Center Constrained Fuzzy C-Means clustering algorithm is proposed. To constrain incorrect migration in the subsequent clustering phase, the algorithm uses a two-stage clustering algorithm. During the first stage weights in the clustering process to improve the accuracy of the resulting clustering centers are introduced while during the second stage, these refined cluster centers can be constrained. Third, a Convolutional Squeeze-and-Excitation Attention Mechanism Wavelet Neural Network (CSWNN) is proposed which more accurately distinguishes uncertain pixels and thus reduces redundant features. The final Denoising Diffusion Probabilistic Model is introduced into the CSWNN to expand pseudolabeled samples. Comparing this algorithm with seven sets of contrast experiments on six real SAR datasets shows that the accuracy of detecting changes in small regions is 2% –9% better than the classical algorithm. © 2008-2012 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Analysis of the Performance of Transformers and Ensemble Schemes for Twitter US Airlines Sentiment Analysis 24

Analysis of the Performance of Transformers and Ensemble Sch...

引用

Proceedings of the 28th Pan-Hellenic Conference on Progress in Computing and informatics

作者： Konstantinos Kyritsis Isidoros Perikos Charalampos M. Liapis Vaggelis Kapoulas Michael Paraskevas Electrical and Computer Engineering University of the Peloponnese Patras Greece Computer Engineering and Informatics Department University of Patras Patras Greece and Computer Technology Institute and Press Diophantus Patras Greece Department of Mathematics University of Patras Patras Greece and Computer Technology Institute and Press Diophantus Patras Greece Computer Technology Institute and Press Diophantus Patras Greece Electrical and Computer Engineering University of the Peloponnese Patras Greece and Computer Technology Institute and Press Diophantus Patras Greece

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：