检索结果-内蒙古大学图书馆

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

Learning Group-Disentangled Representation for Interpretable Thoracic Pathologic Prediction

Learning Group-Disentangled Representation for Interpretable...

引用

2022 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2022

作者： Li, Hao Wu, Yirui Hu, Hexuan Lu, Hu Lai, Yong Wan, Shaohua Hohai University Key Laboratory of Water Big Data Technology of Ministry of Water Resources China College of Computer and Information Hohai University China Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education Jilin University China School of Computer Science and Communication Engineering Jiangsu University China Shenzhen Institute for Advanced Study University of Electronic Science and Technology of China China

ISBN: (纸本)9781665468190

Deep learning methods have shown significant performance in medical image analysis tasks. However, they generally act like 'black box' without explanations in both feature extraction and decision processes, leading to lack of clinical insights and high risk assessments. To aid deep learning in envisioning diseases with visual clues, we propose Representation Group-Disentangling Network (RGD-Net), which can completely disentangle feature space of input X-ray images into several independent feature groups, each corresponding to a specific disease. Taking several semantically related and labeled X-ray images as input, RGD-Net firstly extracts completely group-disentangled representations of diseases through Group-Disentangle Module, which applies group-swap and linking operations to construct latent space by enforcing semantic consistency of attributes. To prevent learning degenerate representations defined as shortcut problem, we further introduce adversarial constricts on mapping from features to diseases, thus avoiding model collapse with former free-form disentanglement. Experiments on chestxray-14 and ChestXpert datasets demonstrate that RGD-Net are effective in predicting diseases with remarkable advantages, which leverage potential factors contributing to different diseases, thus enhancing interpretability in working patterns of deep learning methods. © 2022 IEEE.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

NegativePrompt: Leveraging Psychology for Large Language Models Enhancement via Negative Emotional Stimuli

arXiv

引用

arXiv 2024年

作者： Wang, Xu Li, Cheng Chang, Yi Wang, Jindong Wu, Yuan School of Artificial Intelligence Jilin University China Institute of Software CAS China Microsoft Research Asia China Key Laboratory of Symbolic Computation and Knowledge Engineering Jilin University China International Center of Future Science Jilin University China

Large Language Models (LLMs) have become integral to a wide spectrum of applications, ranging from traditional computing tasks to advanced artificial intelligence (AI) applications. This widespread adoption has spurred extensive research into LLMs across various disciplines, including the social sciences. Notably, studies have revealed that LLMs possess emotional intelligence, which can be further developed through positive emotional stimuli. This discovery raises an intriguing question: can negative emotions similarly influence LLMs, potentially enhancing their performance? In response to this question, we introduce NegativePrompt, a novel approach underpinned by psychological principles, involving ten specifically designed negative emotional stimuli. We embark on rigorous experimental evaluations of five LLMs including Flan-T5-Large, Vicuna, Llama 2, ChatGPT, and GPT-4, across a set of 45 tasks. The results are revealing: NegativePrompt markedly enhances the performance of LLMs, evidenced by relative improvements of 12.89% in Instruction Induction tasks and 46.25% in BIG-Bench tasks. Moreover, we conduct attention visualization experiments to decipher the underlying mechanisms of NegativePrompt’s influence. Our research contributes significantly to the understanding of LLMs and emotion interaction, demonstrating the practical efficacy of NegativePrompt as an emotion-driven method and offering novel insights for the enhancement of LLMs in real-world applications. The code is available at https://***/wangxu0820/NegativePrompt. Copyright © 2024, The Authors. All rights reserved.

关键词： computational linguistics

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

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

Closed Loop Networks for Open-Set Semi-Supervised Learning

SSRN

引用

SSRN 2023年

作者： Ouyang, Jihong Meng, Qingyi Li, Ximing Zhang, Zhengjie Li, Changchun Wang, Wenting College of Computer Science and Technology Jilin University China Key Laboratory of Symbolic Computation and Knowledge Engineering Ministry of Education Jilin University China Department of Computer Science University of Texas Dallas United States

Open-Set Semi-Supervised Learning (OS-SSL) refers to the task of learning classifiers with labeled and unlabeled instances, but the unlabeled data may contain the instances associated with unseen labels, dubbed as Out-Of-Distribution (OOD) instances. The main idea of OS-SSL is to detect and filter out the OOD instances to avoid the negative effects for classifier training. The current methods treat the detection of OOD instances and the classification of seen labels as two specific tasks, and achieve OS-SSL by using a shared backbone with two task-specific heads. However, we found that such way may result in a conflict problem between the two specific heads by observing the gradients of the shared backbone parameters. To solve this problem, we develop a novel OS-SSL method, namely Closed LOop NEtworks (CLONE), which employs two respective networks to detect OOD instances and classify seen labels, but makes a feedback loop between them. We can train CLONE in an end-to-end manner efficiently. We empirically compare CLONE with the existing OS-SSL methods on several benchmark datasets. The experimental results indicate that CLONE can outperform existing baseline methods in most settings. © 2023, The Authors. All rights reserved.

关键词： Cloning

来源：评论

学校读者我要写书评

暂无评论

Todo: Task Offloading Decision Optimizer for the Efficient Provision of Offloading Schemes

SSRN

引用

SSRN 2023年

作者： Chen, Shilin Wang, Xingwang Sun, Yafeng College of Computer Science and Technology Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education Jilin University Changchun130012 China College of Computer Science and Technology Jilin University Changchun130012 China

As the volume of data stored on local devices increases, users turn to edge devices to help with processing tasks. Developing offloading schemes is challenging due to the varying configurations of edge devices and user preferences. While traditional methods provide schemes for offloading in various scenarios, they face unavoidable challenges, including the requirement to manage device workloads in real-time, significant computational costs, and the difficulty of balancing multi-objectives in offloading schemes. To solve these problems, we propose the Task Offloading Decision Optimizer, which offers efficient multi-objective offloading schemes considering real-time device workload and user preference. The proposed offloading scheme contains three goals: reducing task execution time, decreasing device energy consumption, and lowering rental costs. It comprises two important parts: Scheme Maker and Scheme Assistor. Scheme Maker utilizes deep reinforcement learning, optimizes the internal architecture, and enhances the performance of the operation. It optimizes buffer storage to generate dependable multi-objective offloading schemes considering real-time environmental conditions. Scheme Assistor utilizes the data in the Scheme Maker buffer to enhance efficiency by reducing computational costs. Extensive experiments have proved that the proposed framework efficiently provides offloading schemes considering the real-time conditions of the devices and the users, and it offers offloading schemes that enhance task completion rate by 50%. Compared to the baseline, the task execution time is reduced by 12.1%, and the device energy consumption is reduced by 11.4%. © 2023, The Authors. All rights reserved.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Task-Oriented Multi-Modal Mutual Learning for Vision-Language Models

Task-Oriented Multi-Modal Mutual Learning for Vision-Languag...

引用

International Conference on Computer Vision (ICCV)

作者： Sifan Long Zhen Zhao Junkun Yuan Zichang Tan Jiangjiang Liu Luping Zhou Shengsheng Wang Jingdong Wang College of Computer Science and Technology Jilin University Jilin China Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education Jilin University Jilin China Baidu VIS University of Sydney Zhejiang University

Prompt learning has become one of the most efficient paradigms for adapting large pre-trained vision-language models to downstream tasks. Current state-of-the-art methods, like CoOp and ProDA, tend to adopt soft prompts to learn an appropriate prompt for each specific task. Recent CoCoOp further boosts the base-to-new generalization performance via an image-conditional prompt. However, it directly fuses identical image semantics to prompts of different labels and significantly weakens the discrimination among different classes as shown in our experiments. Motivated by this observation, we first propose a class-aware text prompt (CTP) to enrich generated prompts with label-related image information. Unlike CoCoOp, CTP can effectively involve image semantics and avoid introducing extra ambiguities into different prompts. On the other hand, instead of reserving the complete image representations, we propose text-guided feature tuning (TFT) to make the image branch attend to class-related representation. A contrastive loss is employed to align such augmented text and image representations on downstream tasks. In this way, the image-to-text CTP and text-to-image TFT can be mutually promoted to enhance the adaptation of VLMs for downstream tasks. Extensive experiments demonstrate that our method outperforms the existing methods by a significant margin. Especially, compared to CoCoOp, we achieve an average improvement of 4.03% on new classes and 3.19% on harmonic-mean over eleven classification benchmarks.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Learning with Partial Labels from Semi-supervised Perspective

arXiv

引用

arXiv 2022年

作者： Li, Ximing Jiang, Yuanzhi Li, Changchun Wang, Yiyuan Ouyang, Jihong College of Computer Science and Technology Jilin University China Key Laboratory of Symbolic Computation and Knowledge Engineering of MOE Jilin University China College of Information Science and Technology Northeast Normal University China Key Laboratory of Applied Statistics of MOE Northeast Normal University China

Partial Label (PL) learning refers to the task of learning from the partially labeled data, where each training instance is ambiguously equipped with a set of candidate labels but only one is valid. Advances in the recent deep PL learning literature have shown that the deep learning paradigms, e.g., self-training, contrastive learning, or class activate values, can achieve promising performance. Inspired by the impressive success of deep Semi-Supervised (SS) learning, we transform the PL learning problem into the SS learning problem, and propose a novel PL learning method, namely Partial Label learning with Semi-supervised Perspective (PLSP). Specifically, we first form the pseudo-labeled dataset by selecting a small number of reliable pseudo-labeled instances with high-confidence prediction scores and treating the remaining instances as pseudo-unlabeled ones. Then we design a SS learning objective, consisting of a supervised loss for pseudo-labeled instances and a semantic consistency regularization for pseudo-unlabeled instances. We further introduce a complementary regularization for those non-candidate labels to constrain the model predictions on them to be as small as possible. Empirical results demonstrate that PLSP significantly outperforms the existing PL baseline methods, especially on high ambiguity levels. Code available: https://***/changchunli/PLSP. Copyright © 2022, The Authors. All rights reserved.

关键词： Deep learning

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：