检索结果-内蒙古大学图书馆

Federated Offline Reinforcement Learning with Proximal Policy Evaluation

Chinese Journal of Electronics 2024年第6期33卷 1360-1372页

作者： Sheng YUE Yongheng DENG Guanbo WANG Ju REN Yaoxue ZHANG Department of Computer Science and Technology BNRist Tsinghua University Zhongguancun Laboratory

Offline reinforcement learning(RL) has gathered increasing attention in recent years, which seeks to learn policies from static datasets without active online exploration. However, the existing offline RL approaches often require a large amount of pre-collected data and hence are hardly implemented by a single agent in practice. Inspired by the advancement of federated learning(FL), this paper studies federated offline reinforcement learning(FORL),whereby multiple agents collaboratively carry out offline policy learning with no need to share their raw ***, a straightforward solution is to simply retrofit the off-the-shelf offline RL methods for FL, whereas such an approach easily overfits individual datasets during local updating, leading to instability and subpar performance. To overcome this challenge, we propose a new FORL algorithm, named model-free(MF)-FORL, that exploits novel“proximal local policy evaluation” to judiciously push up action values beyond local data support, enabling agents to capture the individual information without forgetting the aggregated knowledge. Further, we introduce a model-based variant, MB-FORL, capable of improving the generalization ability and computational efficiency via utilizing a learned dynamics model. We evaluate the proposed algorithms on a suite of complex and high-dimensional offline RL benchmarks, and the results demonstrate significant performance gains over the baselines.

关键词： Federated learning Computational modeling Heuristic algorithms Reinforcement learning Performance gain Benchmark testing Data models Computational efficiency Trajectory Servers

来源：评论

学校读者我要写书评

暂无评论

Acquiring Weak Annotations for Tumor Localization in Temporal and Volumetric Data

引用

Machine Intelligence Research 2024年第2期21卷 318-330页

作者： Yu-Cheng Chou Bowen Li Deng-Ping Fan Alan Yuille Zongwei Zhou Department of Computer Science Johns Hopkins UniversityBaltimore 21218USA Computer Vision Lab ETH ZurichZurich 8001Switzerland

Creating large-scale and well-annotated datasets to train AI algorithms is crucial for automated tumor detection and ***,with limited resources,it is challenging to determine the best type of annotations when annotating massive amounts of unlabeled *** address this issue,we focus on polyps in colonoscopy videos and pancreatic tumors in abdominal CT scans;Both applications require significant effort and time for pixel-wise annotation due to the high dimensional nature of the data,involving either temporary or spatial *** this paper,we develop a new annotation strategy,termed Drag&Drop,which simplifies the annotation process to drag and *** annotation strategy is more efficient,particularly for temporal and volumetric imaging,than other types of weak annotations,such as per-pixel,bounding boxes,scribbles,ellipses and ***,to exploit our Drag&Drop annotations,we develop a novel weakly supervised learning method based on the watershed *** results show that our method achieves better detection and localization performance than alternative weak annotations and,more importantly,achieves similar performance to that trained on detailed per-pixel ***,we find that,with limited resources,allocating weak annotations from a diverse patient population can foster models more robust to unseen images than allocating per-pixel annotations for a small set of *** summary,this research proposes an efficient annotation strategy for tumor detection and localization that is less accurate than per-pixel annotations but useful for creating large-scale datasets for screening tumors in various medical modalities.

关键词： Weak annotation detection localization segmentation colonoscopy abdomen

来源：评论

学校读者我要写书评

暂无评论

A hybrid feature selection method for text classification using a feature-correlation-based genetic algorithm

引用

Soft Computing 2024年第23期28卷 13567-13593页

作者： Farek, Lazhar Benaidja, Amira Computer Science Department University of Guelma Guelma Algeria Computer Science Department University of Setif 1 Setif Algeria Laboratory of Vision and Artificial Intelligence - LAVIA Larbi Tebessi University Tebessa Algeria

This paper introduces a new hybrid method to address the issue of redundant and irrelevant features selected by filter-based methods for text classification. The method utilizes an enhanced genetic algorithm called "Feature Correlation-based Genetic Algorithm" (FC-GA). Initially, a feature subset with the highest classification accuracy is selected by a filter-based method, which will be then used by the FC-GA to generate potential solutions by considering the correlation between features that have similar classification weights and avoiding useless random solutions. The encoding process involves assigning a value of 0 to features that provide a high degree of correlation with other features having almost the same classification information beyond a specified context, while features that are lowly correlated retain their initial code of 1. Through iterative optimization using crossover and mutation operators, the algorithm should remove redundant features that provide strong correlations and high redundancy, which could lead to improved classification performance at a lower computation cost. The aim of this study is to improve the efficiency of filter-based methods, incorporate feature correlation information into genetic algorithms, and utilize pre-optimized feature subsets to efficiently identify optimal solutions. To evaluate the effectiveness of the proposed method, SVM and NB classifiers are employed on six public datasets and compared to five well-known and effective filter-based methods. The results indicate that a significant portion (about 50%) of the features selected by reference filter-based methods are redundant. Eliminating those redundant features leads to a significant improvement in classification performance as measured by the micro-f1 measure. © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2024.

关键词： Genetic algorithms

来源：评论

学校读者我要写书评

暂无评论

DNACoder: a CNN-LSTM attention-based network for genomic sequence data compression

引用

Neural Computing and Applications 2024年第29期36卷 18363-18376页

作者： Sheena, K.S. Nair, Madhu S. Artificial Intelligence & amp Computer Vision Lab Department of Computer Science Cochin University of Science and Technology Kerala Kochi682022 India

Genomic sequencing has become increasingly prevalent, generating massive amounts of data and facing a significant challenge in long-term storage and transmission. A solution that reduces the storage and transfer requirements without compromising data integrity is needed. The effectiveness of neural networks has already been endorsed in tasks like image and speech compression. Adapting them to recognize the intricate patterns in genomic sequences could help to find more redundancies and reduce storage requirements. The proposed method, called DNACoder, leverages deep learning techniques to achieve significant compression ratios while preserving the essential information in genomic data and offers a high-performance compression for genomic sequences in any data format. The results of the experiments clearly demonstrate the effectiveness of the method and its potential applications in genomic data storage. Our proposed method improves compression by 21.1% on bits per base compared to existing compressors on the benchmarked dataset. By using a deep learning prediction model that is structured as a convolutional layer followed by an attention-based long short-term memory network, we propose a novel lossless and reference-free compression approach (DNACoder), which can also be utilized as a reference-based compressor. The experimental outcome on the tested data illustrates that the advocated compression algorithm’s CNN-LSTM model makes generalizations effectively for genomic sequence data and outperforms the state-of-the-art methods. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2024.

关键词： Long short-term memory

来源：评论

学校读者我要写书评

暂无评论

TextFormer: A Query-based End-to-end Text Spotter with Mixed Supervision

引用

Machine Intelligence Research 2024年第4期21卷 704-717页

作者： Yukun Zhai Xiaoqiang Zhang Xiameng Qin Sanyuan Zhao Xingping Dong Jianbing Shen School of Computer Science Beijing Institute of TechnologyBeijing 100081China Department of Computer Vision Technology Baidu Inc.Beijing 100193China

End-to-end text spotting is a vital computer vision task that aims to integrate scene text detection and recognition into a unified *** methods heavily rely on region-of-interest(Rol)operations to extract local features and complex post-processing steps to produce final *** address these limitations,we propose TextFormer,a query-based end-to-end text spotter with a transformer ***,using query embedding per text instance,TextFormer builds upon an image encoder and a text decoder to learn a joint semantic understanding for multitask *** allows for mutual training and optimization of classification,segmentation and recognition branches,resulting in deeper feature sharing without sacrificing flexibility or ***,we design an adaptive global aggregation(AGG)module to transfer global features into sequential features for reading arbitrarilyshaped texts,which overcomes the suboptimization problem of Rol ***,potential corpus information is utilized from weak annotations to full labels through mixed supervision,further improving text detection and end-to-end text spotting *** experiments on various bilingual(i.e.,English and Chinese)benchmarks demonstrate the superiority of our *** on the TDA-ReCTS dataset,TextFormer surpasses the state-of-the-art method in terms of 1-NED by 13.2%.

关键词： End-to-end text spotting arbitrarily-shaped texts transformer mixed supervision multitask modeling.

来源：评论

学校读者我要写书评

暂无评论

A secret sharing scheme based on integer decomposition and hexagonal structure

引用

International Journal of Information and Communication Technology 2024年第4期24卷 482-501页

作者： Rouia, Zender Lemnouar, Noui Rida, Abdessemed Mohamed LAMIE Laboratory Department of Computer Science Faculty of Mathematics and Computer Science University of Batna 2 Algeria

Security is a major challenge in storage and transmission of digital data. Secret sharing scheme is a fundamental primitive used in multiparty computations, access control and key management, which is based here on two concepts, namely: hexagonal structure and integer decomposition. Use of hexagonal structure is common in biological modelling. For integer decomposition, the oldest known method is Fermat’s factorisation, while for the proposed decomposition, the factorisation uniqueness of positive integer into two factors is exploited. Experimental results obtained from the applied scheme to digital images reveal interesting properties;this scheme turns out to be lossless, ideal, flexible, extensible, and even can detect and identify cheater;in sum, it has a good security. © 2024 Inderscience Enterprises Ltd.

关键词： Factorization

来源：评论

学校读者我要写书评

暂无评论

An infrastructure software perspective toward computation offloading between executable specifications and foundation models

引用

science China(Information sciences) 2025年第4期68卷 380-382页

作者： Dezhi RAN Mengzhou WU Yuan CAO Assaf MARRON David HAREL Tao XIE Key Laboratory of High Confidence Software Technologies (PKU) Ministry of Education School of Computer SciencePeking University School of Electronics Engineering and Computer Science Peking University Department of Computer Science and Applied Mathematics Weizmann Institute of Science

Foundation models(FMs) [1] have revolutionized software development and become the core components of large software systems. This paradigm shift, however, demands fundamental re-imagining of software engineering theories and methodologies [2]. Instead of replacing existing software modules implemented by symbolic logic, incorporating FMs' capabilities to build software systems requires entirely new modules that leverage the unique capabilities of ***, while FMs excel at handling uncertainty, recognizing patterns, and processing unstructured data, we need new engineering theories that support the paradigm shift from explicitly programming and maintaining user-defined symbolic logic to creating rich, expressive requirements that FMs can accurately perceive and implement.

关键词：

来源：评论

学校读者我要写书评

暂无评论

GPS: graph contrastive learning via multi-scale augmented views from adversarial pooling

引用

science China(Information sciences) 2025年第1期68卷 145-158页

作者： Wei JU Yiyang GU Zhengyang MAO Ziyue QIAO Yifang QIN Xiao LUO Hui XIONG Ming ZHANG School of Computer Science National Key Laboratory for Multimedia Information ProcessingPeking University Artificial Intelligence Thrust The Hong Kong University of Science and Technology Department of Computer Science University of California

Self-supervised graph representation learning has recently shown considerable promise in a range of fields, including bioinformatics and social networks. A large number of graph contrastive learning approaches have shown promising performance for representation learning on graphs, which train models by maximizing agreement between original graphs and their augmented views(i.e., positive views). Unfortunately, these methods usually involve pre-defined augmentation strategies based on the knowledge of human experts. Moreover, these strategies may fail to generate challenging positive views to provide sufficient supervision signals. In this paper, we present a novel approach named graph pooling contrast(GPS) to address these *** by the fact that graph pooling can adaptively coarsen the graph with the removal of redundancy, we rethink graph pooling and leverage it to automatically generate multi-scale positive views with varying emphasis on providing challenging positives and preserving semantics, i.e., strongly-augmented view and weakly-augmented view. Then, we incorporate both views into a joint contrastive learning framework with similarity learning and consistency learning, where our pooling module is adversarially trained with respect to the encoder for adversarial robustness. Experiments on twelve datasets on both graph classification and transfer learning tasks verify the superiority of the proposed method over its counterparts.

关键词： graph representation learning graph neural networks graph contrastive learning graph augmentations graph pooling

来源：评论

学校读者我要写书评

暂无评论

Deep Gradient Learning for Efficient Camouflaged Object Detection

引用

Machine Intelligence Research 2023年第1期20卷 92-108页

作者： Ge-Peng Ji Deng-Ping Fan Yu-Cheng Chou Dengxin Dai Alexander Liniger Luc Van Gool School of Computer Science Wuhan UniversityWuhan 430072China Computer Vision Laboratory ETH ZürichZürich 8092Switzerland

This paper introduces deep gradient network(DGNet),a novel deep framework that exploits object gradient supervision for camouflaged object detection(COD).It decouples the task into two connected branches,i.e.,a context and a texture *** es-sential connection is the gradient-induced transition,representing a soft grouping between context and texture *** from the simple but efficient framework,DGNet outperforms existing state-of-the-art COD models by a large ***,our efficient version,DGNet-S,runs in real-time(80 fps)and achieves comparable results to the cutting-edge model JCSOD-CVPR21 with only 6.82%*** application results also show that the proposed DGNet performs well in the polyp segmentation,defect detec-tion,and transparent object segmentation *** code will be made available at https://***/GewelsJI/DGNet.

关键词： Camouflaged object detection(COD) object gradient soft grouping efficient model image segmentation

来源：评论

学校读者我要写书评

暂无评论

Re-quantization based binary graph neural networks

引用

science China(Information sciences) 2024年第7期67卷 160-171页

作者： Kai-Lang YAO Wu-Jun LI National Key Laboratory for Novel Software Technology Department of Computer Science and TechnologyNanjing University

Binary neural networks have become a promising research topic due to their advantages of fast inference speed and low energy consumption. However, most existing studies focus on binary convolutional neural networks, while less attention has been paid to binary graph neural networks. A common drawback of existing studies on binary graph neural networks is that they still include lots of inefficient full-precision operations in multiplying three matrices and are therefore not efficient enough. In this paper, we propose a novel method, called re-quantization-based binary graph neural networks(RQBGN), for binarizing graph neural networks. Specifically, re-quantization, a necessary procedure contributing to the further reduction of superfluous inefficient full-precision operations, quantizes the results of multiplication between any two matrices during the process of multiplying three matrices. To address the challenges introduced by requantization, in RQBGN we first study the impact of different computation orders to find an effective one and then introduce a mixture of experts to increase the model capacity. Experiments on five benchmark datasets show that performing re-quantization in different computation orders significantly impacts the performance of binary graph neural network models, and RQBGN can outperform other baselines to achieve state-of-the-art performance.

关键词： graph neural networks binary neural networks mixture of experts computation-efficient algorithms

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：