检索结果-内蒙古大学图书馆

IEEE International Conference on High Performance Computing and Communications (HPCC)

作者： Hongliang Li Dong Xu Zhewen Xu Xiang Li College of Computer Science and Technology Jilin University Changchun China Key Laboratory of Symbolic Computation and Knowledge Engineering of the Ministry of Education Changchun China

Nowadays, data parallelism has been widely applied to train large datasets on distributed deep learning clusters, but it has suffered from costly global parameter updates at batch barriers. Performance imbalance among worker instances, introduced by uneven workload partitioning or biased resource allocation, can cause straggly workers, which can lead to severe impacts on both training speed and result accuracy. This paper studies the issue focusing on the tradeoff between training speed and result accuracy. We propose Cooperate Grouping Parallel (CGP), a hybrid parameter update solution that allows the flexibility of both synchronous and asynchronous update schemes. We introduce a novel Cooperate Worker Grouping Problem (CWGP) that seeks a task grouping configuration that leads to maximum model accuracy and holds customized training speed guarantees. We propose an evolution-based Pareto local searching algorithm to compute efficient worker grouping configurations. Comprehensive evaluation results are presented to demonstrate the effectiveness of CGP under both persistent and fluctuant imbalances. The proposed method alleviates the imbalance impacts without introducing extra adjustment over-heads.

关键词： Training Deep learning computational modeling Focusing Distributed databases Parallel processing Resource management

来源：评论

学校读者我要写书评

暂无评论

ExplSched: Maximizing Deep Learning Cluster Efficiency for Exploratory Jobs

ExplSched: Maximizing Deep Learning Cluster Efficiency for E...

引用

IEEE International Conference on Cluster Computing

作者： Hongliang Li Hairui Zhao Zhewen Xu Xiang Li Haixiao Xu College of Computer Science and Technology Jilin University Changchun China Key Laboratory of Symbolic Computation and Knowledge Engineering of the Ministry of Education Changchun China High Performance Computing Center Jilin University China

Resource management for Deep Learning (DL) clusters is essential for system efficiency and model training quality. Existing schedulers provided by DL frameworks are mostly adaptations from traditional HPC clusters and usually work on jobs’ makespan, assuming that DL training jobs finish completely. Unfortunately, it is reported that a fair amount of training jobs are exploratory jobs and often finish unsuccessfully (over 30%) in production clusters. This is due to the distinct characteristic of Deep Neural Network (DNN) training that it is an exploratory process of frequent user interventions, such as adjusting model structures, tuning hyperparameters, and exploring feature validity. Existing DL cluster schedulers using offline algorithms are not suitable for exploratory jobs when unexpected early terminations can cause noticeable resource waste. Moreover, DL training jobs are iterative and usually yield diminishing returns as they progress. Equally allocating resource among training iterations is not efficient, especially when dealing with exploratory jobs where it can worsen the degradation of system efficiency. The fundamental goal of a DL training job is to gain model quality improvement, usually indicated by the loss reduction (job profit) of a DNN model. This paper introduces a novel scheduling problem for exploratory jobs that seeks to maximize the overall training profit of a DL cluster. We propose ExplSched, an online scheduling solution based on the primal-dual framework, resulting in a competitive ratio of 2α that belongs to O(ln n). It uses a resource price function that emphasizes the importance of job profit to resource consumption ratio to make quick resource allocation decisions. Experimental results show that ExplSched achieved an average system utility improvement of 87.28% compared with other related work.

关键词：

来源：评论

学校读者我要写书评

暂无评论

DISLOCATIONS WITH CORNERS IN AN ELASTIC BODY WITH APPLICATIONS TO FAULT DETECTION

arXiv

引用

arXiv 2023年

作者： Diao, Huaian Liu, Hongyu Meng, Qingle School of Mathematics Key Laboratory of Symbolic Computation and Knowledge Engineering Ministry of Education Jilin University Jilin Changchun China Department of Mathematics City University of Hong Kong Kowloon Tong Hong Kong

This paper focuses on an elastic dislocation problem that is motivated by applications in the geophysical and seismological communities. In our model, the displacement satisfies the Lamé system in a bounded domain with a mixed homogeneous boundary condition. We also allow the occurrence of discontinuities in both the displacement and traction fields on the fault curve/surface. By the variational approach, we first prove the well-posedness of the direct dislocation problem in a rather general setting with the Lamé parameters being real-valued L∞ functions and satisfy the strong convexity condition. Next, by considering that the Lamé parameters are constant and the fault curve/surface possesses certain corner singularities, we establish a local characterization of the jump vectors at the corner points over the dislocation curve/surface. In our study, the dislocation is geometrically rather general and may be open or closed. We establish the unique results for the inverse problem of determining the dislocation curve/surface and the jump vectors for both cases. © 2023, CC BY.

关键词： Inverse problems

来源：评论

学校读者我要写书评

暂无评论

Multi-site MRI classification using Weighted federated learning based on Mixture of Experts domain adaptation

Multi-site MRI classification using Weighted federated learn...

引用

2022 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2022

作者： Bai, Tian Zhang, Yingfang Wang, Yuzhao Qin, Yanguo Zhang, Fa Jilin University College of Computer Science and Technology Changchun China Jilin University Key Laboratory of Symbolic Computation and Knowledge Engineering Ministry of Education Changchun China The Second Hospital of Jilin University The Department of Orthopaedics Changchun China Beijing Institute of Technology School of Medical Technology Beijing China

ISBN: (纸本)9781665468190

Deep learning often requires large amounts of data from different institutions. Federated learning, as a distributed training framework, enables multiple participants to collaboratively train models without collecting data together and hence protecting data privacy, but the datasets from different institutions usually bring the problem of domain shift, which affects the performance of the model. When addressing domain shift, previous works often use a single global model to share parameters. Therefore, we propose a novel method to train multiple public models with different structures under the federated framework to improve the reliability and robustness of the public models. And each participant keeps its own domain-tuned private model, the private model does not share parameters with other participants. We use Mixture of Experts (MoE) domain adaptation to dynamically combine different public models and private model, which utilizes the similarity between different datasets to update the parameters of the public models. We apply the proposed method to the multi-site Magnetic resonance imaging (MRI) end-to-end classification, and the experiments demonstrate its effectiveness. © 2022 IEEE.

关键词： Magnetic resonance imaging

来源：评论

学校读者我要写书评

暂无评论

Hybrid features and semantic reinforcement network for image forgery detection

Hybrid features and semantic reinforcement network for image...

引用

作者： Chen, Haipeng Chang, Chaoqun Shi, Zenan Lyu, Yingda College of Computer Science and Technology Jilin University Changchun130012 China College of Software Jilin University Changchun130012 China Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education Jilin University Changchun130012 China Public Computer Education and Research Center Jilin University Changchun130012 China

Image forgery detection focuses more on tampering regions than image content of semantic segmentation, it is revealed that wealthier features need to be learned. Moreover, insufficient semantic information causes low efficiency of forgery detection. To address these issues, we propose a hybrid features and semantic reinforcement network (HFSRNet) for image forgery detection, which is an encoding and decoding based network. Specifically, long-short term memory with resampling features has been applied to capture traces from the image patches for finding manipulating artifacts. Consolidated features extracted from rotating residual units are further leveraged to amplify the discrepancy between un-tampered and tampered regions. We then hybridize features from them through a concatenation to further incorporate spatial co-occurrence of these two modalities. In addition, for achieving the semantic consistency between two same level features associated by across layers, semantic reinforcement is implemented on the decoding stage. HFSRNet is an end-to-end architecture that handles multiple types of image forgery including copy-move, splicing, removal. Experiments on three standard image manipulation datasets (NIST16, COVERAGE and CASIA) demonstrate that HFSRNet obtains state-of-the-art performance compared to existing models and baselines. © 2021, The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

3DSEAVNet: 3D-Squeeze-and-Excitation Networks for Audio-Visual Saliency Prediction

3DSEAVNet: 3D-Squeeze-and-Excitation Networks for Audio-Visu...

引用

International Joint Conference on Neural Networks (IJCNN)

作者： Silong Liang Chunxiao Li Naying Cui Minghui Sun Hao Xue College of Software Engineering JiLin University Changchun China College of Computer Science and Technology JiLin University Changchun China Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education JiLin University Changchun China

Video saliency prediction is an important task in the field of computer vision. Most of the existing video saliency prediction methods only focus on image information, and the audio information is often ignored. This leads to an incomplete perception mode, which makes it difficult to achieve optimal performance. SENet is an excellent attention mechanism-based network. It significantly enhances the performance of 2D convolutional networks. However, whether the 3D convolutional network can be applied to this attention mechanism network remains to be studied. In order to solve the above problems, we propose a saliency prediction network for audio-visual fusion to extract and predict various information in videos. At the same time, we improve the traditional SENet to make it applicable in 3D convolutional neural networks and discuss its role. Compared with the state-of-the-art methods, our model has strong competitiveness in multiple data sets.

关键词：

来源：评论

学校读者我要写书评

暂无评论

RFR-WWANet: Weighted Window Attention-Based Recovery Feature Resolution Network for Unsupervised Image Registration

arXiv

引用

arXiv 2023年

作者： Ma, Mingrui Wang, Tao Wang, Weijie Song, Lei Liu, Guixia College of Computer Science and Technology Jilin University Changchun China Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education Changchun China Department of Information Engineering and Computer Science University of Trento Trento Italy

The Swin transformer has recently attracted attention in medical image analysis due to its computational efficiency and long-range modeling capability. Owing to these properties, the Swin Transformer is suitable for establishing more distant relationships between corresponding voxels in different positions in complex abdominal image registration tasks. However, the registration models based on transformers combine multiple voxels into a single semantic token. This merging process limits the transformers to model and generate coarse-grained spatial information. To address this issue, we propose Recovery Feature Resolution Network (RFRNet), which allows the transformer to contribute fine-grained spatial information and rich semantic correspondences to higher resolution levels. Furthermore, shifted window partitioning operations are inflexible, indicating that they cannot perceive the semantic information over uncertain distances and automatically bridge the global connections between windows. Therefore, we present a Weighted Window Attention (WWA) to build global interactions between windows automatically. It is implemented after the regular and cyclic shift window partitioning operations within the Swin transformer block. The proposed unsupervised deformable image registration model, named RFR-WWANet, detects the long-range correlations, and facilitates meaningful semantic relevance of anatomical structures. Qualitative and quantitative results show that RFR-WWANet achieves significant improvements over the current state-of-the-art methods. Ablation experiments demonstrate the effectiveness of the RFRNet and WWA designs. Our code is available at https://***/MingR-Ma/RFR-WWANet. Copyright © 2023, The Authors. All rights reserved.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

A Re-Parametrization-Based Bayesian Differential Analysis Algorithm for Gene Regulatory Networks Modeled with Structural Equation Models

引用

Computer Modeling in engineering & Sciences 2020年第7期124卷 303-313页

作者： Yan Li Dayou Liu Yungang Zhu Jie Liu College of Computer Science and Technology Jilin UniversityChangchun130012China Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education Jilin UniversityChangchun130012China

Under different conditions,gene regulatory networks(GRNs)of the same gene set could be similar but *** differential analysis of GRNs under different conditions is important for understanding condition-specific gene regulatory *** a naive approach,existing GRN inference algorithms can be used to separately estimate two GRNs under different conditions and identify the differences between ***,in this way,the similarities between the pairwise GRNs are not taken into *** joint differential analysis algorithms have been proposed recently,which were proved to outperform the naive approach *** this paper,we model the GRNs under different conditions with structural equation models(SEMs)to integrate gene expression data and genetic perturbations,and re-parameterize the pairwise SEMs to form an integrated model that incorporates the differential ***,a Bayesian inference method is used to make joint differential analysis by solving the integrated *** evaluated the performance of the proposed re-parametrization-based Bayesian differential analysis(ReBDA)algorithm by running simulations on synthetic data with different *** performance of the ReBDA algorithm was demonstrated better than another state-of-the-art joint differential analysis algorithm for SEMs ReDNet *** the end,the ReBDA algorithm was applied to make differential analysis on a real human lung gene data set to illustrate its applicability and practicability.

关键词： Gene regulatory networks structural equation models joint

来源：评论

学校读者我要写书评

暂无评论

Causality-inspired Unsupervised Domain Adaptation with Target Style Imitation for Medical Image Segmentation

引用

IEEE Transactions on Circuits and Systems for Video Technology 2025年

作者： Song, Jincai Chen, Haipeng Lyu, Yingda Nie, Weizhi Liu, An-An Jilin University College of Computer Science and Technology Jilin Changchun130012 China Jilin University Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education Jilin Changchun130012 China Jilin University Public Computer Education and Research Center Jilin Changchun130012 China Tianjin University School of Electrical and Information Engineering Tianjin300072 China

Deep learning performance may decrease substantially with unseen heterogeneous data. While most unsupervised domain adaptation (UDA) methods seek to address this through image alignment, they often ignore uncertainty style fluctuations within the target domain. When testing image styles vary in both direction and intensity, such models may fail to adapt. Furthermore, existing UDA methods tend to over-reliance on domain-level entire feature alignment, resulting in potentially over-exploiting semantic content-independent cues (e.g., intensity) as shortcut features. To address these limitations, this paper introduces an innovative and model-agnostic Causality-inspired Representation Learning Based on Target Style Imitation method for UDA. Specifically, we propose a novel Target Style Imitation (TSI) data augmentation approach to diversify the training data and align training and unseen target testing image styles. TSI constructs a Gaussian distribution for the target domain style and simulates unseen testing style variations through random sampling. Additionally, inspired by the stable and generalizable causal mechanism, we propose Causality-inspired Representation Learning (CRL) based on TSI method to enforce feature representations to adhere to causal properties (i.e., Separation and Independence) essential for robust UDA, thus fostering the model to focus on the domain-invariant semantic features. Our method surpasses state-of-the-art methods on two cross-modality medical image segmentation datasets. © 1991-2012 IEEE.

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

Scalable Precise computation of Shannon Entropy

arXiv

引用

arXiv 2025年

作者： Lai, Yong Tong, Haolong Xu, Zhenghang Yin, Minghao College of Computer Science and Technology Jilin University Changchun130012 China Key Laboratory of Symbolic Computation and Knowledge Engineering Ministry of Education Jilin University Changchun130012 China School of Computer Science and Information Technology Northeast Normal University Changchun130017 China

Quantitative information flow analyses (QIF) are a class of techniques for measuring the amount of confidential information leaked by a program to its public outputs. Shannon entropy is an important method to quantify the amount of leakage in QIF. This paper focuses on the programs modeled in Boolean constraints and optimizes the two stages of the Shannon entropy computation to implement a scalable precise tool PSE. In the first stage, we design a knowledge compilation language called ADD[∧] that combines Algebraic Decision Diagrams and conjunctive decomposition. ADD[∧] avoids enumerating possible outputs of a program and supports tractable entropy computation. In the second stage, we optimize the model counting queries that are used to compute the probabilities of outputs. We compare PSE with the state-of-the-art probably approximately correct tool EntropyEstimation, which was shown to significantly outperform the existing precise tools. The experimental results demonstrate that PSE solved 55 more benchmarks compared to EntropyEstimation in a total of 441. For 98% of the benchmarks that both PSE and EntropyEstimation solved, PSE is at least 10× as efficient as EntropyEstimation. Copyright © 2025, The Authors. All rights reserved.

关键词： Information leakage

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：