检索结果-内蒙古大学图书馆

Imbalanced multilabel retinal disease classification using threshold moving and ensemble learning

Multimedia Tools and Applications 2025年 1-19页

作者： Pendharkar, Gaurav Balaji, Sudharshanan Kumar, B. Muhesh Malathi, G. School of Computer Science and Engineering Vellore Institute of Technology Tamil Nadu Chennai600127 India

Billions of people worldwide are affected by vision impairment majorly caused due to age-related degradation and refractive errors. Diabetic Retinopathy(DR) and Macular Hole(MH) are among the most prevalent senescent retinal diseases. Machine Intelligence can assist ophthalmologists and clinicians in fast and accurate disease diagnosis by identifying patterns in disease progression for a better healthcare system. In this paper, the Retinal Fundus Multi-disease Image Dataset (RFMiD) is used to design a machine intelligence system with two chief components namely a disease risk classifier, and a multi-label classifier. The disease risk classifier predicts whether the retinal fundus image is infected or not. Based on the prediction of the disease risk classifier, a multi-label classifier can be applied to obtain probabilities for the susceptibility to DR and MH. Finally, an ensemble is employed with the best of 3 models for each classifier. The disease riskpredictor attained a peak F1-score of 88%, while the multi-label classifier achieved an Area Under the Curve(AUC) score of 86%. However, the individual binary classifiers for DR and MH reached maximum F-scores of 91% and 93%, respectively. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2025.

关键词： Risk perception

来源：评论

学校读者我要写书评

暂无评论

A Generative Image Steganography Based on Disentangled Attribute Feature Transformation and Invertible Mapping Rule

引用

computers, Materials & Continua 2025年第4期83卷 1149-1171页

作者： Xiang Zhang Shenyan Han Wenbin Huang Daoyong Fu School of Computer Science Nanjing University of Information Science and TechnologyNanjing210044China Engineering Research Center of Digital Forensics Nanjing University of Information Science and TechnologyMinistry of EducationNanjing210044China

Generative image steganography is a technique that directly generates stego images from secret *** traditional methods,it theoretically resists steganalysis because there is no cover ***,the existing generative image steganography methods generally have good steganography performance,but there is still potential room for enhancing both the quality of stego images and the accuracy of secret information ***,this paper proposes a generative image steganography algorithm based on attribute feature transformation and invertible mapping ***,the reference image is disentangled by a content and an attribute encoder to obtain content features and attribute features,***,a mean mapping rule is introduced to map the binary secret information into a noise vector,conforming to the distribution of attribute *** noise vector is input into the generator to produce the attribute transformed stego image with the content feature of the reference ***,we design an adversarial loss,a reconstruction loss,and an image diversity loss to train the proposed *** results demonstrate that the stego images generated by the proposed method are of high quality,with an average extraction accuracy of 99.4%for the hidden ***,since the stego image has a uniform distribution similar to the attribute-transformed image without secret information,it effectively resists both subjective and objective steganalysis.

关键词： Image information hiding generative information hiding disentangled attribute feature transformation invertible mapping rule steganalysis resistance

来源：评论

学校读者我要写书评

暂无评论

Simulation Study of Rayleigh-Bénard Thermal Convection Based on PINN 7

Simulation Study of Rayleigh-Bénard Thermal Convection Base...

引用

7th International Conference on computer Information science and Application technology, CISAT 2024

作者： Zou, Qingfeng He, Jie Liu, Tao Xie, Jianan Zhao, Dongmei University of Science and Technology School of Computer Science and Technology Mianyang China

ISBN: (纸本)9798350375107

In order to overcome the limitations of traditional numerical methods in the numerical simulation of two-dimensional Rayleigh-Bénard heat convection, a novel approach based on physical information neural networks (PINNs) is proposed. This method addresses the challenges of high mesh cost, low computational efficiency, and complex boundary condition processing. The thermal convection field is computed for four Rayleigh numbers 1.1 × 106} ≤ Ra ≤ 1.1 × 109) at Pr=0.72. In this work, the thermal convection field data solved by COMSOL® software is used to construct the data set for the PINNs model with the size of 8040000, and the data set is divided into a training set ([0, 8000000]) and a test set ([8000000, 8040000]), and the predictions were made for the test dataset using the trained model and validated by comparing the predictions with the thermal convection field computed with COMSOL® for the same operating conditions. As an illustrative example, consider the case where 1.1 × 106 is used. The results demonstrate that the maximum relative errors of the temperature and flow fields solved by the model are 1.28 × 10-3 and 3.16 × 10-3, respectively. The accuracy of the model prediction of the PINNs can be maintained at around 10-2, although it is found that the accuracy of the model prediction of the PINNs slightly decreases with the larger Ra number. Moreover, the model prediction requires approximately 5 seconds, whereas the COMSOL® computation time is 73 seconds. In comparison, the computational efficiency of the PINNs model is improved by a factor of 14. The results demonstrate that PINNs can not only effectively simulate the 2D Rayleigh-Bénard thermal convection problem, but also significantly reduce its numerical simulation time. © 2024 IEEE.

关键词： Rayleigh number

来源：评论

学校读者我要写书评

暂无评论

Dual-Task Contrastive Meta-Learning for Few-Shot Cross-Domain Emotion Recognition

引用

computers, Materials & Continua 2025年第2期82卷 2331-2352页

作者： Yujiao Tang Yadong Wu Yuanmei He Jilin Liu Weihan Zhang School of Computer Science and Engineering Sichuan University of Science and EngineeringYibin644002China School of Mechanical and Power Engineering Chongqing University of Science and TechnologyChongqing401331China

Emotion recognition plays a crucial role in various fields and is a key task in natural language processing (NLP). The objective is to identify and interpret emotional expressions in text. However, traditional emotion recognition approaches often struggle in few-shot cross-domain scenarios due to their limited capacity to generalize semantic features across different domains. Additionally, these methods face challenges in accurately capturing complex emotional states, particularly those that are subtle or implicit. To overcome these limitations, we introduce a novel approach called Dual-Task Contrastive Meta-Learning (DTCML). This method combines meta-learning and contrastive learning to improve emotion recognition. Meta-learning enhances the model’s ability to generalize to new emotional tasks, while instance contrastive learning further refines the model by distinguishing unique features within each category, enabling it to better differentiate complex emotional expressions. Prototype contrastive learning, in turn, helps the model address the semantic complexity of emotions across different domains, enabling the model to learn fine-grained emotions expression. By leveraging dual tasks, DTCML learns from two domains simultaneously, the model is encouraged to learn more diverse and generalizable emotions features, thereby improving its cross-domain adaptability and robustness, and enhancing its generalization ability. We evaluated the performance of DTCML across four cross-domain settings, and the results show that our method outperforms the best baseline by 5.88%, 12.04%, 8.49%, and 8.40% in terms of accuracy.

关键词： Contrastive learning emotion recognition cross-domain learning dual-task meta-learning

来源：评论

学校读者我要写书评

暂无评论

Online Stable Streaming Feature Selection via Feature Aggregation

引用

ACM Transactions on Knowledge Discovery from Data 2025年第3期19卷 1-23页

作者： Zhou, Peng Wang, Qi Zhang, Yunyun Ling, Zhaolong Zhao, Shu Wu, Xindong School of Computer Science and Technology Anhui University Hefei China Anhui University Hefei China Hefei University of Technology Hefei China

Feature selection is an essential pre-process component in data mining that aims to select the most relevant features from the target dataset. Datasets are always dynamic in real-world applications, and features may exist in stream mode. Then, online streaming feature selection methods are proposed, which deal with streaming features arriving continuously in real-time. However, most existing algorithms prioritize high accuracy and low time-consumption but overlook the stability of the selected features. Stable feature selection results are crucial for users in practice. For instance, in the medical field, unstable feature selection results can make it challenging for experts to identify the main causative factors of a disease. Motivated by this, this article proposes a new online stable streaming feature selection method via feature aggregation named OSSFS. Specifically, inspired by the cohesive MeanShift approach, OSSFS applies an incremental aggregation strategy to partition the streaming features into multiple hyperellipsoids. Then, we incrementally update and merge these hyperellipsoids with new streaming features. Finally, we select representative features from each hyperellipsoid as the final selected feature subset. Extensive experiments are conducted on several real-world datasets to compare our new method with state-of-the-art competing algorithms in cases of stability and predictive accuracy. Experimental results indicate that OSSFS achieves optimal stability without losing prediction accuracy. © 2025 Copyright is held by the owner/author(s). Publication rights licensed to ACM.

关键词： Feature Selection Feature Streams Stable Feature Selection Feature Aggregation

来源：评论

学校读者我要写书评

暂无评论

PBCounter:weighted model counting on pseudo-boolean formulas

引用

Frontiers of computer science 2025年第3期19卷 55-63页

作者： Yong LAI Zhenghang XU Minghao YIN College of Computer Science and Technology Jilin UniversityChangchun 130012China Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education Jilin UniversityChangchun 130012China School of Computer Science and Information Technology Northeast Normal UniversityChangchun 130017China

In Weighted Model Counting(WMC),we assign weights to literals and compute the sum of the weights of the models of a given propositional formula where the weight of an assignment is the product of the weights of its *** current WMC solvers work on Conjunctive Normal Form(CNF)***,CNF is not a natural representation for human-being in many *** by the stronger expressive power of Pseudo-Boolean(PB)formulas than CNF,we propose to perform WMC on PB *** on a recent dynamic programming algorithm framework called ADDMC for WMC,we implement a weighted PB counting tool *** compare PBCounter with the state-of-the-art weighted model counters SharpSAT-TD,ExactMC,D4,and ADDMC,where the latter tools work on CNF with encoding methods that convert PB constraints into a CNF *** experiments on three domains of benchmarks show that PBCounter is superior to the model counters on CNF formulas.

关键词： weighted model counting pseudo-boolean constraint algebraic decision diagrams preprocessing techniques

来源：评论

学校读者我要写书评

暂无评论

Hybrid Beamforming Considering Path Birth-death Process via Few-shot Learning 17

Hybrid Beamforming Considering Path Birth-death Process via ...

引用

17th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics, CISP-BMEI 2024

作者： Huang, Ruixi Long, Yin Luo, Qiaomu Liu, Yu School of Computer Science and Technology Southwest University of Science and Technology China

ISBN: (纸本)9798331507398

While data-driven approaches provide an effective solution for hybrid beamforming implementation, they currently fail to account for the path birth-death process (PBP), a phenomenon frequently encountered in practical environments. The occurrence of path birth-death process (PBP) leads to abrupt and significant shifts in the statistical properties of channel state information, which, in turn, necessitates frequent model updates to adjust to the altered channel conditions. Consequently, models generated by existing data-driven techniques that primarily focus on angle range variation are inadequate when path birth-death process (PBP) is present. In this paper, we propose a weight initialization strategy for hybrid beamforming model design, inspired by few-shot learning, to mitigate the challenges posed by path birth-death process (PBP). With this approach, the model can efficiently adapt to changes in the number of propagation channels by relying on starting weights and requiring only minimal online updates. Simulation outcomes show that the introduced technique greatly enhances adaptation efficiency compared to previous methods. © 2024 IEEE.

关键词： Channel state information

来源：评论

学校读者我要写书评

暂无评论

Timely Data Delivery for Energy-Harvesting IoT Devices

引用

Chinese Journal of Electronics 2022年第2期31卷 322-336页

作者： LU Wenwei GONG Siliang ZHU Yihua School of Computer Science and Technology Zhejiang University of Technology College of Mathematics and Computer Science Zhejiang A&F University

The devices in the Internet of things(Io T) gain capability of sustainable operation when they harvest energy from ambient sources. Fluctuation in the harvested energy may cause the energy-harvesting IoT devices to suffer from frequent energy shortage, which may bring in intolerable packet delay or packet discarding. It is important to design a low-delay packet delivery scheme that adapts to variation in the harvested *** this paper, we present the timely data delivery(TDD)scheme for the IoT devices. Using Markov chain, we develop a probability model for the TDD scheme, which leads to the expected number of packets delivered in an operation cycle, the expected numbers of packets waiting in the data buffer in an operation cycle and an energy-harvesting cycle, and the expected packet delay. Additionally, we formulate the optimization problem that minimizes the packet delay in the TDD scheme, and the solution to the optimization problem yields the optimal parameters for the IoT devices to determine when to harvest energy and when to deliver data under the TDD *** simulation results show that the proposed TDD scheme outperforms the existing schemes in terms of packet delay.

关键词： Internet of things energy-harvesting IoT devices expected packet delay Markov processes Markov chain probability frequent energy shortage low-delay packet delivery scheme operation cycle delays harvested energy intolerable packet delay optimisation energy-harvesting cycle telecommunication power management energy harvesting timely data delivery scheme TDD scheme packet discarding probability model Internet of Things

来源：评论

学校读者我要写书评

暂无评论

Graph Attention Networks Fusing Semantic Information for Knowledge Graph Completion 17

Graph Attention Networks Fusing Semantic Information for Kno...

引用

17th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics, CISP-BMEI 2024

作者： Li, Min Li, Xuejun Liao, Jing School of Computer Science and Technology Southwest University of Science and Technology China

ISBN: (纸本)9798331507398

Knowledge graph embedding aims to embed triples into low-dimensional vector spaces. Through the attention mechanism, models based on the graph attention network can learn the information of the neighboring node, improving knowledge graph completion and enriching the structural information. On the other hand, because of the shallow graph attention network, these models can only learn local structural information and cannot capture the global semantic properties of entities and relations. They also only use linear matrices to update relations, ignoring the impact of neighboring information on relations. We propose a fused semantic information graph attention network (FSI-GAT) to address the above challenges for knowledge graph completion. FSI-GAT consists of three modules: structural information learning module, semantic information learning module, and scoring module. The structural information learning module adopts the improved graph attention network to learn the local structural information of entities and relations;the semantic information learning module encodes the descriptive statements of entities and relations through the BERT pretrained language model to obtain the global semantic information;and the scoring module uses the ConvE to score the candidate tail entities. The experimental results show that FSI-GAT significantly outperforms the existing models on the link prediction task and can effectively improve the link prediction performance. © 2024 IEEE.

关键词： Knowledge graph

来源：评论

学校读者我要写书评

暂无评论

SA-Model:Multi-Feature Fusion Poetic Sentiment Analysis Based on a Hybrid Word Vector Model

引用

computer Modeling in Engineering & sciences 2023年第10期137卷 631-645页

作者： Lingli Zhang Yadong Wu Qikai Chu Pan Li Guijuan Wang Weihan Zhang Yu Qiu Yi Li School of Computer Science and Engineering Sichuan University of Science and EngineeringZigong643000China School of Automation and Information Engineering Sichuan University of Science and EngineeringZigong643000China School of Computer Science and Technology Southwest University of Science and TechnologyMianyang621000China School of Information Engineering Southwest University of Science and TechnologyMianyang621000China

Sentiment analysis in Chinese classical poetry has become a prominent topic in historical and cultural tracing,ancient literature research,***,the existing research on sentiment analysis is relatively *** does not effectively solve the problems such as the weak feature extraction ability of poetry text,which leads to the low performance of the model on sentiment analysis for Chinese classical *** this research,we offer the SA-Model,a poetic sentiment analysis ***-Model firstly extracts text vector information and fuses it through Bidirectional encoder representation from transformers-Whole word masking-extension(BERT-wwmext)and Enhanced representation through knowledge integration(ERNIE)to enrich text vector information;Secondly,it incorporates numerous encoders to remove text features at multiple levels,thereby increasing text feature information,improving text semantics accuracy,and enhancing the model’s learning and generalization capabilities;finally,multi-feature fusion poetry sentiment analysis model is *** feasibility and accuracy of the model are validated through the ancient poetry sentiment *** with other baseline models,the experimental findings indicate that SA-Model may increase the accuracy of text semantics and hence improve the capability of poetry sentiment analysis.

关键词： Sentiment analysis Chinese classical poetry natural language processing BERT-wwm-ext ERNIE multi-feature fusion

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：