检索结果-内蒙古大学图书馆

Proximal recursive generalized hyper-gradient descent method

APPLIED SOFT COMPUTING 2025年 175卷

作者： Zhang, Hao Lu, Shuxia Hebei Univ Coll Math & Informat Sci Key Lab Machine Learning & Computat Intelligence H Baoding 071002 Peoples R China

This paper focuses on the non-convex, non-smooth composite optimization problem. It consists of a non-convex loss function and a non-smooth regularizer function that admits a proximal mapping. However, the method is still limited in handling objective functions that involve non-smooth regularizer. How to determine the step size for solving composite optimization problems can be a challenge. To address this gap, we propose a recursive gradient descent algorithm using generalized hyper-gradient descent, named ProxSarah-GHD, which utilizes variance reduction techniques and provides update rules for adaptive step sizes. To improve its generalization in proximal gradient descent, a generalized variant of hyper-gradient descent, named Generalized Hyper-gradient Descent (GHD), is proposed in this paper. We prove that ProxSarah-GHD attains a linear convergence rate. Moreover, we provide the oracle complexity of ProxSarah-GHD as O (epsilon-3) and O (\/n epsilon-2 + n) in the online setting and finite-sum setting, respectively. In addition, to avoid the trouble of manually adjusting the batch size, we develop a novel Exponentially Increasing Mini-batch scheme for ProxSarah-GHD, named ProxSarah-GHD-EIM. The theoretical analysis that shows ProxSarah-GHD-EIM achieves a linear convergence rate is also provided, and shows that its total complexity is O (epsilon-4 + epsilon-2) and O(n + epsilon-4 + epsilon-2) in the online setting and finite-sum setting, respectively. Numerical experiments on standard datasets verify the superiority of the ProxSarah-GHD over other methods. We further analyze the sensitivity of the ProxSarah-GHD-EIM to its hyperparameters, conducting experiments on standard datasets.

关键词： Non-convex optimization Generalized hyper-gradient descent Stochastic recursive gradient algorithm Adaptive step size

来源：评论

学校读者我要写书评

暂无评论

Global Well-Posedness for the 2D Keller-Segel-Navier-Stokes System with Fractional Diffusion

引用

ACTA APPLICANDAE MATHEMATICAE 2024年第1期194卷 7页

作者： Wang, Chaoyong Jia, Qi Zhang, Qian Hebei Univ Coll Math & Informat Sci Key Lab Machine Learning & Computat Intelligence Baoding 071002 Peoples R China

In this paper, we consider Cauchy problem for the 2D incompressible Keller-Segel-Navier-Stokes equations with the fractional diffusion {partial derivative tn+u & sdot;del n-Delta n=-del & sdot;(n del c)-n3,partial derivative tc+u & sdot;del c-Delta c=-c+n,partial derivative tu+u & sdot;del u+boolean AND 2 alpha u+del P=-n del phi,\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document} $$\begin{aligned} \left \{ \begin{aligned} &\partial _{t}n+u\cdot \nabla n-\Delta n=-\nabla \cdot (n\nabla c)- n<^>{3}, \\ &\partial _{t}c+u\cdot \nabla c-\Delta c=-c+n, \\ &\partial _{t}u+u\cdot \nabla u+\wedge <^>{2\alpha }u+\nabla P=-n\nabla \phi , \end{aligned} \right . \end{aligned}$$ \end{document} where boolean AND:=(-Delta)12\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$\wedge :=(-\Delta )<^>{\frac{1}{2}}$\end{document} and alpha is an element of[12,1]\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$\alpha \in [\frac{1}{2},1]$\end{document}. We get the global well-posedness for the above system with the rough initial data by a new priori estimate of the solutions.

关键词： Keller-Segel equations Generalized Navier-Stokes equations Global well-posedness

来源：评论

学校读者我要写书评

暂无评论

Resonance response and chaotic analysis for an irrational pendulum system

引用

CHAOS SOLITONS & FRACTALS 2024年 182卷

作者： Han, Ning Zhang, Hanfang Lu, Peipei Liu, Zixuan Hebei Univ Coll Math & Informat Sci Key Lab Machine Learning & Computat Intelligence Baoding 071002 Peoples R China

We propose a novel rotating pendulum system with irrational nonlinearity, bi-stable phenomenon and quasizero stiffness characteristic, which bears a transition from standard double well dynamics to discontinuous double well dynamics by adjusting a geometric parameter. Nonlinear dynamical behaviors of the proposed irrational pendulum system subjected to viscus damping and periodic excitation are studied. To analytically demonstrate the primary resonance response for this perturbed irrational pendulum system, we introduce an approximate irrational system which is significant parallels with the original pendulum system from the perspective of qualitative analysis and quantitative calculation. Averaging method is applied to investigate the dynamic response of the perturbed irrational pendulum system with quasi-zero stiffness at origin. The effect of the internal and external parameters on the response curves are discussed. Numerical fitting technique and semi-analytical Melnikov method are used to detect the chaotic boundaries of the perturbed irrational pendulum system with two types of homoclinic orbits. Different types of periodic solutions and chaos are clarified in the perturbed irrational pendulum system by using numerical simulations. It is found that a class of chaos with oscillatory and rotational motions bears significant similarities to paroxysmal chaos due to sudden rotational motion and two chaotic motions formed by period-doubling bifurcation merge into one chaos.

关键词： Irrational pendulum Dynamic response Chaotic boundary Averaging method Melnikov method Periodic solution

来源：评论

学校读者我要写书评

暂无评论

LSPC: Exploring contrastive clustering based on local semantic information and prototype

引用

INFORMATION SYSTEMS 2024年 121卷

作者： Chen, Jun-Fen Sun, Lang Xie, Bo-Jun Hebei Univ Coll Math & Informat Sci Key Lab Machine Learning & Computat Intelligence Baoding 071002 Hebei Peoples R China

Recently years, several prominent contrastive learning algorithms, a kind of self-supervised learning methods, have been extensively studied that can efficiently extract useful feature representations from input images by means of data augmentation techniques. How to further partition the representations into meaningful clusters is the issue that deep clustering is addressing. In this work, a deep clustering algorithm based on local semantic information and prototype is proposed referring to LSPC that aims at learning a group of representative prototypes. Rather than learning the distinguishing characteristics between different images, more attention is given to the essential characteristics of images that are maybe from a potential category. On the training framework, contrastive learning is skillfully combined with k-means clustering algorithm. The prediction is transformed into soft assignments for end-to-end training. In order to enable the model to accurately capture the semantic information between images, we mine similar samples of training samples in the embedded space as local semantic information to effectively increase the similarity between samples belonging to the same cluster. Experimental results show that our algorithm achieves state-of-the-art performance on several commonly used public datasets, and additional experiments prove that this superior clustering performance can also be extended to large datasets such as ImageNet.

关键词： Image clustering Unsupervised learning Contrastive learning Self-supervised learning

来源：评论

学校读者我要写书评

暂无评论

Big data decision tree for continuous-valued attributes based on unbalanced cut points

引用

JOURNAL OF BIG DATA 2023年第1期10卷 135页

作者： Ma, Shixiang Zhai, Junhai Hebei Univ Coll Math & Informat Sci Key Lab Machine Learning & Computat Intelligence Baoding Peoples R China

The decision tree is a widely used decision support model, which can quickly mine effective decision rules based on the dataset. The decision tree induction algorithm for continuous-valued attributes, based on unbalanced cut points, is efficient for mining decision rules;however, extending it to big data remains an unresolved. In this paper, two solutions are proposed to solve this problem: the first one is based on partitioning instance subsets, whereas the second one uses partitioning attribute subsets. The crucial of these two solutions is how to find the global optimal cut point from the set of local optimal cut points. For the first solution, the calculation of the Gini index of the cut points between computing nodes and the selection of the global optimal cut point by communication between these computing nodes is proposed. However, in the second solution, the division of the big data into subsets using attribute subsets in a way that all cut points of an attribute are on the same map node is proposed, the local optimal cut points can be found in this map node, then the global optimal cut point can be obtained by summarizing all local optimal cut points in the reduce node. Finally, the proposed solutions are implemented with two big data platforms, Hadoop and Spark, and compared with three related algorithms on four datasets. Experimental results show that the proposed algorithms can not only effectively solve the scalability problem, but also have lowest running time, the fastest speed and the highest efficiency under the premise of preserving the classification performance.

关键词： Big data Decision tree Cut point Hadoop Spark

来源：评论

学校读者我要写书评

暂无评论

A self-supervised learning of semantic feature consistency for image clustering

引用

JOURNAL OF INTELLIGENT & FUZZY SYSTEMS 2023年第5期45卷 8651-8661页

作者： Chen, Junfen Han, Jie Xie, Bojun Li, Nana Hebei Univ Coll Math & Informat Sci Key Lab Machine Learning & Computat Intelligence Baoding Peoples R China

Contrastive learning is a powerful technique for learning feature representations without manual annotation. The K-nearest neighbor (KNN) method is commonly used to construct positive sample pairs to calculate the contrastive loss. However, it is challenging to distinguish positive sample pairs, reducing clustering performance. We propose a novel Deep Contrastive Clustering method based on a GrapH convolutional network called GHDCC. It uses an instance-level contrastive loss with mean square error (MSE) regularization and a cluster-level contrastive loss to incorporate semantic features and perform cluster assignments. The method utilizes a graph convolutional network (GCN) to improve the semantic consistency of features and linear interpolation data augmentation to improve the representation ability of the model. To minimize the occurrence of false positive sample pairs, we select only samples whose similarity exceeds a predefined threshold to construct the adjacency matrix. The experimental results on six public datasets demonstrate that the GHDCC significantly outperforms contrastive clustering (CC, 500) by a large margin except on CIFAR-10. The GHDCC performs well compared to other deep contrastive clustering methods and achieves the highest clustering accuracy of 0.913 on ImageNet-10.

关键词： Self-supervised clustering graph convolutional network linear interpolation data augmentation contrastive learning

来源：评论

学校读者我要写书评

暂无评论

Unknown Support Prototype Set for Open Set Recognition

引用

INTERNATIONAL JOURNAL OF COMPUTER VISION 2025年第7期133卷 4366-4383页

作者： Jiang, Guosong Zhu, Pengfei Cao, Bing Chen, Dongyue Hu, Qinghua Tianjin Univ Coll Intelligence & Comp Tianjin Key Lab Machine Learning Tianjin 300350 Peoples R China

In real-world applications, visual recognition systems inevitably encounter unknown classes which are not present in the training set. Open set recognition aims to classify samples from known classes and detect unknowns, simultaneously. One promising solution is to inject unknowns into training sets, and significant progress has been made on how to build an unknowns generator. However, what unknowns exhibit strong generalization is rarely explored. This work presents a new concept called Unknown Support Prototypes, which serve as good representatives for potential unknown classes. Two novel metrics coined Support and Diversity are introduced to construct Unknown Support Prototype Set. In the algorithm, we further propose to construct Unknown Support Prototypes in the semantic subspace of the feature space, which can largely reduce the cardinality of Unknown Support Prototype Set and enhance the reliability of unknowns generation. Extensive experiments on several benchmark datasets demonstrate the proposed algorithm offers effective generalization for unknowns.

关键词： Open set recognition Data augmentation Unknown detection Prototype

来源：评论

学校读者我要写书评

暂无评论

Dynamic pricing and inventory model for perishable products with reference price and reference quality

引用

JOURNAL OF computatIONAL AND APPLIED MATHEMATICS 2025年 458卷

作者： Chen, Xuexue You, Cuilian Hebei Univ Coll Math & Informat Sci Baoding 071002 Peoples R China Hebei Univ Key Lab Machine Learning & Computat Intelligence Baoding 071002 Peoples R China

Reference price (quality) is a benchmark point used by consumers to make price (quality) judgements. Combining reference price and reference quality with dynamic pricing and inventory management, this paper applies differential equations theory to construct a optimal control model for perishable products when the quality of products declines exponentially. The demand of products depends on price, quality, reference price and reference quality, among which reference price and reference quality are influenced by consumers' past memory. The aim is to maximize the retailer's profit during the period. The conclusions are as follows. Firstly, a optimal dynamic pricing and inventory model for perishable products with reference price and reference quality is constructed, and the model is extended to dual decision variables, stochastic demand, time-dependent effects, competitive and infinite planing horizon setups. Secondly, through the Pontryagin maximum principle, the analytical expression of the optimal dynamic price is derived. Thirdly, a linear search method to solve the optimal static price is proposed. Finally, the sensitivity of the main parameters is analyzed and the corresponding management enlightenments are given. By comparison of dynamic and static pricing, we find that dynamic pricing can achieve more profits and take a shorter selling period. In addition, for the sales problem of perishable products affected by both reference price and reference quality, retailers should adopt skimming pricing strategy (the optimal price decreases with time). Furthermore, to obtain more profit, retails should strive to increase the sensitivity coefficients of two types of reference, and the reference quality memory coefficient, while to decrease the reference price memory coefficient.

关键词： Reference price Reference quality Maximum principle Pricing Inventory

来源：评论

学校读者我要写书评

暂无评论

Unsupervised cross-domain object detection based on dynamic smooth cross entropy

引用

INTERNATIONAL JOURNAL OF machine learning AND CYBERNETICS 2025年 1-14页

作者： Xie, Bojun Huang, Zhijin Chen, Junfen Key Lab Machine Learning & Computat Intelligence Baoding 071002 Hebei Peoples R China Hebei Univ Coll Math & Informat Sci Baoding 071002 Hebei Peoples R China

Pseudo-label self-training methods are highly regarded in cross-domain object detection (CDOD) models for their reduced annotation costs, strong applicability, and straightforward deployment. However, existing methods still struggle to balance the quantity and quality of pseudo-labels, as well as their efficient utilization. To tackle these challenges, this study introduces a novel approach by integrating strong-weak data augmentation (SWDA) techniques into the decoupled adaptation framework, significantly enhancing the generalization ability and accuracy of the pseudo-label generator. Furthermore, inspired by Softmatch, this research proposes a dynamic smooth cross-entropy loss function that adjusts pseudo-label confidence thresholds in real-time based on the model's learning state and the predictive distribution across categories. This method not only retains low-confidence pseudo-labels but also assigns varied weights to them according to their confidence levels, enhancing both the quality and utility of pseudo-labels. This approach substantially improves CDOD performance, especially demonstrated in the Pascal VOC ->\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\rightarrow$$\end{document} Clipart task, where it outperformed the leading Adaptive Teacher and Class-Aware Teacher methods by 2.3 and 2.5% points, respectively, achieving an accuracy of 51.6%. The method also demonstrated commendable performance in other commonly used unsupervised CDOD tasks. The code is available at https://***/985767019/DSCE.

关键词： Decoupled adaptation Strong-weak data augmentation Dynamic smooth cross entropy Pseudo-labels Cross-domain object detection

来源：评论

学校读者我要写书评

暂无评论

Joint Correcting and Refinement for Balanced Low-Light Image Enhancement

引用

IEEE TRANSACTIONS ON MULTIMEDIA 2024年 26卷 6310-6324页

作者： Yu, Nana Shi, Hong Han, Yahong Tianjin Univ Coll Intelligence & Comp Tianjin Key Lab Machine Learning Tianjin 300072 Peoples R China

Low-light image enhancement tasks demand an appropriate balance among brightness, color, and illumination. While existing methods often focus on one aspect of the image without considering how to pay attention to this balance, which will cause problems of color distortion and overexposure etc. This seriously affects both human visual perception and the performance of high-level visual models. In this work, a novel synergistic structure is proposed which can balance brightness, color, and illumination more effectively. Specifically, the proposed method, so-called Joint Correcting and Refinement Network (JCRNet), which mainly consists of three stages to balance brightness, color, and illumination of enhancement. Stage 1: we utilize a basic encoder-decoder and local supervision mechanism to extract local information and more comprehensive details for enhancement. Stage 2: cross-stage feature transmission and spatial feature transformation further facilitate color correction and feature refinement. Stage 3: we employ a dynamic illumination adjustment approach to embed residuals between predicted and ground truth images into the model, adaptively adjusting illumination balance. Extensive experiments demonstrate that the proposed method exhibits comprehensive performance advantages over 21 state-of-the-art methods on 9 benchmark datasets. Furthermore, a more persuasive experiment has been conducted to validate our approach the effectiveness in downstream visual tasks (e.g., saliency detection). Compared to several enhancement models, the proposed method effectively improves the segmentation results and quantitative metrics of saliency detection.

关键词： Image color analysis Lighting Feature extraction Brightness Image enhancement Distortion Visualization Low-light enhancement back projection color correction illumination adjustment

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：