检索结果-内蒙古大学图书馆

Hybrid TCN-transformer model for predicting sustainable food supply and ensuring resilience

Alexandria engineering Journal 2025年 127卷 380-391页

作者： Alrashdi, Ibrahim Abd El-Aziz, Rasha M. Taloba, Ahmed I. Farsi, Mohammed Department of Computer Science College of Computer and Information Sciences Jouf University Al-Jouf Sakaka73211 Saudi Arabia Department of Information Systems College of Computer Science and Engineering Taibah University Yanbu966144 Saudi Arabia

Major concerns occur in maintaining a sustainable food supply due to population expansion, supply chain interruptions, and climate-related changes. Traditional forecasting models, such as ARIMA, LSTM, and GRU, fail to deal with the dynamic fluctuations and long-range correlations. These methods often have reduced accuracy, high computational cost, and poor generalization when applied to shifting patterns in the food supply system. Based on the abovementioned limitations, this study proposes a Hybrid TCN-Transformer Model by combining the strength of Temporal Convolutional Networks (TCN) with Transformer Attention to provide accurate and efficient food supply prediction. Thus, the project's key objective is to develop a deep learning system that effectively handles massive quantities of temporal data while preserving dependence in short and long durations. The proposed method is novel since it integrates TCN's causal convolutions for extracting sequential patterns. It uses a Transformer Model's self-attention mechanism to capture the complex interactions across time steps. Hybrid design enables faster training, increased interpretability, and better prediction accuracy than current methods. Results from experiments have revealed that the suggested model surpasses the performance of the stand-alone TCN, ARIMA, LSTM, and GRU models in terms of accuracy of predictions, efficiency of computations, and adaptability. Findings thus far suggest deep learning-based predictive analytics has helped improve the food supply chain management by mitigating food wastage, making it environmentally resilient. © 2025 The Authors

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Detection of Stuttered Speech with Repetition Using Wavelet Packet Transformation and Neuro Fuzzy Classifier

引用

SN computer science 2025年第5期6卷 1-9页

作者： Drakshayini, K.B. Anusuya, M.A. Vani, H.Y. Chayadevi, M.L. Department of Computer Science and Engineering VTU Belagaum India Department of Computer Science and Engineering JSSSTU Mysuru India Department of Information Science and Engineering JSSSTU Mysuru India Department of Computer Science and Engineering BNMIT Mysuru India

Repetition is the most prominent type of dysfluency in stuttered speech. Spectral energy, entropy, ZCR, and centroid are extracted using wavelet packet transformation, and similarity between these features for adjacent frames is measured linguistic hedges on these factors are fed as input to fuzzy ANN. Fuzzy ANN is used to classify stuttering speech signals with repetition and fluent speech signals. Correcting repetition in stuttered speech is handled by fuzzy ANN based on fuzzy rules passed concerning similarity measures of spectral parameters like spectral energy, ZCR, entropy, and centroid. The same process was also examined for fuzzy entropy alone for analysis purposes for different kinds of repetition in stuttered speech. © The Author(s), under exclusive licence to Springer Nature Singapore Pte Ltd. 2025.

关键词： Adaptive neuro fuzzy classifier Centroid Energy Fuzzy entropy Linguistic hedges Wavelet packet transformation ZCR

来源：评论

学校读者我要写书评

暂无评论

An ensembled-deep-learning paradigm trained with a self-improved coyote optimization algorithm (SI-COA) for crop disease detection

引用

Multimedia Tools and Applications 2025年第4期84卷 1697-1724页

作者： Shukla, Preeti Chandanan, Amit Kumar Department of Computer Science and Information Technology Guru Ghasidas Vishwavidyalaya CG Bilaspur India Department of Computer Science & Engineering Guru Ghasidas Vishwavidyalaya CG Bilaspur India

In recent times, drastic climate changes have caused a substantial increase in the growth of crop diseases. This causes large-scale demolition of crops, decreases cultivation, and eventually leads to the financial loss of farmers. Due to the rapid growth in a variety of diseases and adequate knowledge of farmers, identification, and treatment of the disease have become a major challenge. The leaves have texture and visual similarities which are attributes for the identification of disease type. Hence, computer vision employed with deep learning provides a way to solve this problem. In this research work, a novel voting-based ensembled-deep-learning paradigm is developed for crop disease detection. Initially, the collected raw images are pre-processed via Laplacian Filter and morphological operations. From, the pre-processed image, the ROI region is identified using K-means Clustering. Then, from the ROI regions, the features like texture feature (Chi-squared Local Binary Pattern (CLBP)), Scale Invariant Feature Transform (SIFT), Gray-Level Co-occurrence Matrix (GLCM)), color features, and shape features are extracted. Then, the optimal features are selected by using Self-Improved Coyote Optimization Algorithm (SI-COA) is an extended version of the standard Coyote Optimization Algorithm (COA) and is a metaheuristic for optimization that is inspired by the canis latrans species. For crop disease detection, the voting-based ensembled-deep-learning paradigm was used which includes the Attention-based Bi-LSTM, Recurrent Neural Networks (RNNs), and Optimized Deep Neural Networks (O-DNN). The weight of DNN is optimized via Self-Improved Coyote Optimization Algorithm (SI-COA). The overall outcome of this process is evaluated by using various performance metrics such as Precision, Recall, F1-Score, Accuracy, Sensitivity and Specificity, NPV, FNR, FPR, and MCC, as well. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2

关键词： Long short-term memory

来源：评论

学校读者我要写书评

暂无评论

Deep Learning Based Side-Channel Attack Detection for Mobile Devices Security in 5G Networks

引用

清华大学学报自然科学版（英文版） 2025年第3期30卷 1012-1026页

作者： Amjed A.Ahmed Mohammad Kamrul Hasan Ali Alqahtani Shayla Islam Bishwajeet Pandey Leila Rzayeva Huda Saleh Abbas Azana Hafizah Mohd Aman Nayef Alqahtani Center for Cyber Security Faculty of Information Science and TechnologyUniversiti Kebangsaan MalaysiaBangi 43600Malaysia Imam Alkadhim University College Department of Computer Techniques EngineeringBaghdad 10066Iraq Center for Cyber Security Faculty of Information Science and TechnologyUniversiti Kebangsaan MalaysiaBangi 43600Malaysia Department of Networks and Communications Engineering College of Computer Science and Information SystemsNajran UniversityNajran 61441Saudi Arabia Institute of Computer Science and Digital Innovation UCSI UniversityKuala Lumpur 56000Malaysia Department of Intelligent Systems and Cyber Security Astana IT UniversityAstana 20000Kazakstan Department of Computer Science College of Computer Science and EngineeringTaibah UniversityMadinah 42353Saudi Arabia Department of Electrical Engineering College of EngineeringKing Faisal UniversityAl-Hofuf 31982Saudi Arabia

Mobile devices within Fifth Generation(5G)networks,typically equipped with Android systems,serve as a bridge to connect digital gadgets such as global positioning system,mobile devices,and wireless routers,which are vital in facilitating end-user communication ***,the security of Android systems has been challenged by the sensitive data involved,leading to vulnerabilities in mobile devices used in 5G *** vulnerabilities expose mobile devices to cyber-attacks,primarily resulting from security ***-permission apps in Android can exploit these channels to access sensitive information,including user identities,login credentials,and geolocation *** such attack leverages"zero-permission"sensors like accelerometers and gyroscopes,enabling attackers to gather information about the smartphone's *** underscores the importance of fortifying mobile devices against potential future *** research focuses on a new recurrent neural network prediction model,which has proved highly effective for detecting side-channel attacks in mobile devices in 5G *** conducted state-of-the-art comparative studies to validate our experimental *** results demonstrate that even a small amount of training data can accurately recognize 37.5％of previously unseen user-typed ***,our tap detection mechanism achieves a 92％accuracy rate,a crucial factor for text *** findings have significant practical implications,as they reinforce mobile device security in 5G networks,enhancing user privacy,and data protection.

关键词： Fifth Generation(5G)networks smartphone information leakage Side-Channel Attack(SCA) deep learning

来源：评论

学校读者我要写书评

暂无评论

Automating skin cancer screening: a deep learning

引用

Journal of engineering and Applied science 2025年第1期72卷 1-18页

作者： Rashad, Nada M. Abdelnapi, Noha MM. Seddik, Ahmed F. Sayedelahl, M.A. Department of Computer Science Faculty of Computers and Information Suez University Suez43221 Egypt Biomedical Engineering Department Faculty of Engineering Helwan University Cairo11795 Egypt Department of Computer Science Faculty of Computers and Information Damanhour University Damanhour22511 Egypt

Skin cancer presents in various forms, including squamous cell carcinoma (SCC), basal cell carcinoma (BCC), and melanoma. Established risk factors include ultraviolet (UV) radiation exposure from solar or artificial sources, lighter skin pigmentation, a history of sunburns, and a family history of the disease. Early detection and prompt intervention are crucial for achieving a favorable prognosis. Traditionally, treatment modalities include surgery, radiation therapy, and chemotherapy. Recent advancements in immunotherapy have revolutionized skin cancer diagnosis, but manual identification remains time-consuming. Artificial intelligence (AI) has shown potential in skin cancer classification, leading to automated screening methods. To support dermatologists, we improved the model for classifying images. This model is able to recognize seven different kinds of skin lesions. On the ISIC dataset, an analysis has been done. This study offers a novel approach to early skin cancer diagnosis based on image processing. Our approach leverages the high accuracy of a specific convolutional neural network architecture, utilizing transfer learning with pre-trained data to further enhance detection performance. Our findings demonstrate that the employed ResNet-50 transfer learning model achieves a remarkable accuracy of 97%, while ResNet50 without augmentation gives an accuracy of 81.57% and an F1-score of 75.75%. © The Author(s) 2025.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Leveraging Concise Concepts with Probabilistic Modeling for Interpretable Visual Recognition

引用

IEEE Transactions on Multimedia 2025年 27卷 3117-3131页

作者： Zhang, Yixuan Liu, Chuanbin Liu, Yizhi Gao, Yifan Lu, Zhiying Xie, Hongtao Zhang, Yongdong University of Science and Technology of China School of Information Science and Technology China Hunan University of Science and Technology Department of Computer Science and Engineering China

Interpretable visual recognition is essential for decision-making in high-stakes situations. Recent advancements have automated the construction of interpretable models by leveraging Visual Language Models (VLMs) and Large Language Models (LLMs) with Concept Bottleneck Models (CBMs), which process a bottleneck layer associated with human-understandable concepts. However, existing methods suffer from two main problems: a) the collected concepts from LLMs could be redundant with task-irrelevant descriptions, resulting in an inferior concept space with potential mismatch. b) VLMs directly map the global deterministic image embeddings with fine-grained concepts results in an ambiguous process with imprecise mapping results. To address the above two issues, we propose a novel solution for CBMs with Concise Concept and Probabilistic Modeling (CCPM) that can achieve superior classification performance via high-quality concepts and precise mapping strategy. Fisrt, we leverage in-context examples as category-related clues to guide LLM concept generation process. To mitigate redundancy in the concept space, we propose a Relation-Aware Selection (RAS) module to obtain a concise concept set that is discriminative and relevant based on image-concept and inter-concept relationships. Second, for precise mapping, we employ a Probabilistic Distribution Adapter (PDA) that estimates the inherent ambiguity of the image embeddings of pre-trained VLMs to capture the complex relationships with concepts. Extensive experiments indicate that our model achieves state-of-the-art results with a 5.48% improvement in classification accuracy on eight mainstream recognition benchmarks as well as reliable explainability through interpretable analysis. © 1999-2012 IEEE.

关键词： Decision making

来源：评论

学校读者我要写书评

暂无评论

An Experimental Study on Dynamic Lifelong Learning with GPT for Mitigating Catastrophic Forgetting in Aspect-Based Sentiment Analysis

引用

IEEE Access 2025年 13卷 90316-90332页

作者： Huang, Huang Asemi, Adeleh Universiti Malaya Faculty of Computer Science Information Technology Department of Software Engineering Kuala Lumpur Malaysia

GPT is widely recognized as one of the most versatile and powerful large language models, excelling across diverse domains. However, its significant computational demands often render it economically unfeasible for individuals and small businesses, underscoring the need for efficient, domainspecific alternatives. In Aspect-Based Sentiment Analysis (ABSA), existing models are typically optimized for single domains, facing challenges in performing effectively across multiple domains. A key issue, known as "catastrophic forgetting," arises when models trained on one domain lose previously learned knowledge upon exposure to new domain data. This leads to two significant problems: limited cross-domain generalization and difficulty in retaining prior knowledge while learning domain-specific information. To address these challenges, we introduce the Dynamic Lifelong Learning Aspect-Based Sentiment Analysis GPT (DllaGPT), a model designed to handle multiple domains while mitigating catastrophic forgetting. Leveraging datasets from four ABSA domains—Laptops, Restaurants, Tweets, and Finance—this study fine-tunes a pretrained GPT model from HuggingFace sequentially across domains. DllaGPT employs a mechanism to retain real data from earlier domains during new domain training, effectively preserving prior knowledge. Experimental results highlight that DllaGPT achieves an average accuracy of 0.85 and a Backward Transfer (BWT) score of -0.09 across the four domains, showcasing its high accuracy and robust lifelong learning capabilities. © 2025 Institute of Electrical and Electronics Engineers Inc.. All rights reserved.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Exploring Nonuniform Structure of NSRTDs for Characterizing MACA Rules in the Null-Boundary Condition

引用

Complex Systems 2025年第1期34卷 91-127页

作者： Hazra, Suvadip Banerjee, Som Dalui, Mamata Chakraborty, Bidesh Department of Computer Science and Engineering Indian Institute of Information Technology Karnataka Dharwad58009 India Department of Computer Science and Engineering National Institute of Technology West Bengal Durgapur713209 India Department of Computer Science and Engineering Haldia Institute of Technology West Bengal Haldia721657 India

The cellular automaton (CA), a discrete model, is gaining popularity in simulations and scientific exploration across various domains, including cryptography, error-correcting codes, VLSI design and test pattern generation. This paper examines a two-state three-neighborhood nonuniform finite CA with a specified number of fixed points. We utilize a graph-based tool known as the next state rule minterm transition diagram (NSRTD) to analyze the spatiotemporal behavior of these cellular automata (CAs) across all lengths. We devise methods for synthesizing nonuniform single length cycle multi-attractor CAs (MACAs), a specialized class of irreversible CA with a predefined number of fixed points for any arbitrary length. Despite the method’s exponential worstcase time complexity, it offers the advantage of selecting rules for each CA cell from a set of candidate rules, ensuring the desired number of fixed points. © 2025, Complex Systems Publications, Inc. All rights reserved.

关键词： cellular automata fixed-point attractor length cycle multi-attractor cellular automata MACA next state rule minterm transition diagram NSRTD

来源：评论

学校读者我要写书评

暂无评论

AMCB: enhancing the authentication process with blockchain integrated with PUF and machine learning

引用

Multimedia Tools and Applications 2025年第8期84卷 4809-4835页

作者： Al-Ghuraybi, Hind A. AlZain, Mohammed A. Soh, Ben Department of Information Technology College of Computers and Information Technology Taif University Taif21944 Saudi Arabia Department of Computer Science and Computer Engineering La Trobe University Bundoora Australia

The rapid advancement of technology has given rise to medical cyber-physical systems (MCPS), a subset of cyber-physical systems (CPS) specifically tailored for patient care and healthcare providers. MCPS generate substantial volumes of big data, posing challenges in terms of processing, storage, and safeguarding against unauthorized access. This paper introduces the Authenticated Medical Cyber-Physical Blockchain (AMCB) model within MCPS, leveraging Blockchain technology, machine learning, and physically unclonable functions (PUF) to fortify the authentication process, particularly for accessing electronic health records (EHR) stored on cloud servers. The model also involves analyzing data from authenticated devices to ensure its integrity. Experiments on the AMCB model evaluate its efficiency across four layers. In the physical layer, PUF keys are extracted using a built Static Random-Access Memory (SRAM) circuit and stored in a simulated Blockchain, yielding a Hamming Distance standard deviation of 6.60%. The inspection layer analyzes data from authenticated devices using machine learning, with the Random Forest achieving the highest accuracy at 99.21%. In the application layer, centered on user authentication, the average transaction time is 36.15 MS. Overall, the AMCB model requires an average of 49.82 ms to execute operations across all layers. These findings underscore the effectiveness of the proposed model in securing EHR in the dynamic landscape of MCPS. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2024.

关键词： Authentication

来源：评论

学校读者我要写书评

暂无评论

Enhancing the efficiency of lung cancer screening: predictive models utilizing deep learning from CT scans

引用

Neural Computing and Applications 2025年 1-19页

作者： Tawfeek, Medhat A. Alrashdi, Ibrahim Alruwaili, Madallah Shaban, Warda M. Talaat, Fatma M. Department of Computer Science Faculty of Computers and Information Menoufia University Shebin Elkom32511 Egypt Department of Computer Science College of Computer and Information Sciences Jouf University Sakakah Saudi Arabia Department of Computer Engineering and Networks College of Computer and Information Sciences Jouf University Sakakah Saudi Arabia Department of Communication and Electronics Engineering Nile Higher Institute for Engineering and Technology Mansoura Egypt Faculty of Artificial Intelligence Kafrelsheikh University Kafrelsheikh Egypt Faculty of Computer Science and Engineering New Mansoura University Gamasa35712 Egypt

Lung cancer is the most lethal form of cancer. This paper introduces a novel framework to discern and classify pulmonary disorders such as pneumonia, tuberculosis, and lung cancer by analyzing conventional X-ray and CT scan images called lung cancer risk prediction (LCRP) model. LCRP has four modules, namely data collection and preprocessing, data augmentation module, image segmentation module, and prediction module. Actually, LCRP employs three deep learning models;sequential model, functional model, and transfer model on publicly available training datasets. Convolutional neural networks (CNNs) have emerged as a highly effective field in machine learning, particularly for image datasets in the field of biomedical applications. The primary goal is to validate these models by comparing their performance with other models in order to determine their effectiveness in addressing challenging datasets. Our research has revealed a noteworthy enhancement in the efficiency of binary and multi-class classification using mask R-CNN image segmentation. During the model training process, a combination of Adam and stochastic gradient descent dual optimizers has been used to improve performance. LCRP have outperformed current pre-trained models by minimizing training parameters, computational costs, and overhead. It introduces 98.5% accuracy, 88.7% specificity, 89% sensitivity, 89.2% precision, and 89.09% F-measure. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2025.

关键词： Lung cancer

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：