检索结果-内蒙古大学图书馆

Neural Combinatorial Optimization Algorithms for Solving Vehicle Routing Problems: A Comprehensive Survey with Perspectives

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Wu, Xuan Wang, Di Wen, Lijie Xiao, Yubin Wu, Chunguo Wu, Yuesong Yu, Chaoyu Maskell, Douglas L. Zhou, You Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education College of Computer Science and Technology Jilin university Changchun130012 China Joint NTU-UBC Research Centre of Excellence in Active Living for the Elderly Nanyang Technological University 639798 Singapore School of Software Tsinghua University Beijing100084 China College of Computing and Data Science Nanyang Technological University 639798 Singapore

Although several surveys on Neural Combinatorial Optimization (NCO) solvers specifically designed to solve Vehicle Routing Problems (VRPs) have been conducted, they did not cover the state-of-the-art (SOTA) NCO solvers emerged recently. More importantly, to establish a comprehensive and up-to-date taxonomy of NCO solvers, we systematically review relevant publications and preprints, categorizing them into four distinct types, namely Learning to Construct, Learning to Improve, Learning to Predict-Once, and Learning to Predict-Multiplicity solvers. Subsequently, we present the inadequacies of the SOTA solvers, including poor generalization, incapability to solve large-scale VRPs, inability to address most types of VRP variants simultaneously, and difficulty in comparing these NCO solvers with the conventional Operations Research algorithms. Simultaneously, we discuss on-going efforts, identify open inadequacies, as well as propose promising and viable directions to overcome these inadequacies. Notably, existing efforts focus on only one or two of these inadequacies, with none attempting to address all of them concurrently. In addition, we compare the performance of representative NCO solvers from the Reinforcement, Supervised, and Unsupervised Learning paradigms across VRPs of varying scales. Finally, following the proposed taxonomy, we provide an accompanying web page as a live repository for NCO solvers. Through this survey and the live repository, we aim to foster further advancements in the NCO community. Copyright © 2024, The Authors. All rights reserved.

关键词： Combinatorial optimization

Stochastic Gradient Descent Optimizer and Multi-Layer Perceptron in Tensorflow for Improved Vehicular Engine Health Monitoring System Decision Model Performance

学校读者我要写书评

暂无评论

SSRN

SSRN 2023年

作者： Rahim, Md Abdur Rahman, Md. Mustafizur Shofiqul, Md Islam Arafatur, Md. Rahman Ramasamy, D. Department of Mechanical Engineering College of Engineering Universiti Malaysia Pahang Lebuhraya Tun Razak Gambang Kuantan Pahang 26300 Malaysia Faculty of Computing Universiti Malaysia Pahang Pahag Kuantan26300 Malaysia Automotive Engineering Centre Universiti Malaysia Pahang Pahang Pekan26600 Malaysia School of Mathematics and Computer Science University of Wolverhampton United Kingdom

Vehicles have transformed modern transportation by offering speed, efficiency, and convenience. However, increased reliance on them has led to challenges that necessitate sustainable solutions. Unexpected and irregular engine issues can result in failures of other components, intermittent maintenance, significant maintenance costs, and reduced engine longevity. Vehicular risk management systems can address these challenges using the Multi-layer Perceptron (MLP) of the Deep Learning-based Vehicular Engine Health Monitoring System (VEHMS) decision model. The MLP for monitoring vehicular engine health, as an intelligent risk management system, exhibits a considerable generalization gap between the learning and validation curves, indicating flaws in the decision model. Minimizing this gap is crucial for ensuring the model's applicability in vehicles and effective engine health monitoring to reduce unwanted expenses. This study successfully achieved the lowest generalization gap by using the TensorFlow library with Leaky ReLU (Rectified Linear Unit) activation function and stochastic gradient descent (SGD) optimizer in the Deep Learning Multi-layer Perceptron Model, rather than Keras with ReLU activation function and ADAM (adaptive moment estimation) optimizer. This approach significantly increased decision accuracy. The TensorFlow-based VEHMS decision model, employing Leaky ReLU activation function and SGD optimizer, is anticipated to precisely diagnose and categorize vehicular engine conditions as Good, Minor, Moderate, or Critical. Consequently, the proposed model could reduce unwanted expenses and extend engine longevity, contributing to improved vehicular risk management systems. In summary, this study has the potential to enhance vehicular engine health monitoring systems, reduce unnecessary costs, and ultimately lead to more efficient and effective transportation systems. © 2023, The Authors. All rights reserved.

关键词： Stochastic systems

Beyond the Black Box: Explainable AI for Glaucoma Detection and Future Improvements

学校读者我要写书评

暂无评论

Beyond the Black Box: Explainable AI for Glaucoma Detection ...

International Conference on computing and Networking Technology (ICCNT)

作者： Avadhani Bindu Senthil Kumar Thangavel K Somasundaram Sathyan Parthasaradhi Ram Gopal Pulgurthi Meenakshi Y Dhar Department of Computer Science and Engineering Amrita School of Computing Coimbatore Amrita Vishwa Vidyapeetham India Department of Mathematics Amrita School of Physical Sciences Coimbatore Amrita Vishwa Vidyapeetham India Sathyan Eye Hospital and Coimbatore Glaucoma Foundation Ramanathapuram Coimbatore India Medivista Eye Centre Safilguda Secunderabad India Department of Ophthalmology Amrita Institute of Medical Sciences Faridabad India

ISBN: (数字)9798350370249

ISBN: (纸本)9798350370270

Timely and precise identification of glaucoma is essential to avert irreversible vision loss. Automating glaucoma detection is essential whilst the disease worsens over time and manual screening methods possess limitations. It facilitates early identification and treatment in order to prevent the vision impairment. Machine learning and deep learning algorithms are efficient, nevertheless, they suffer from a lack of interpretability and require a substantial amount of data and computational resources. Transfer learning proves efficient by improving scarcity of data with enhanced performance. This study seeks to examine the operational characteristics of transfer learning approaches and employ XAI (eXplainable Artificial Intelligence) method LIME to achieve clear interpretability of the features obtained by the model for prediction, so as to address the challenges related to interpretability. This study has used a combined dataset consisting of RIMONE, PAPILA, CHAKSU, and AKSHI (a dataset created by the authors) to develop ResNet-50, ResNet50-V2, VGG-16, VGG-19, and EfficientNet-V2S models. Among the developed five models, ResNet-50 and EfficientNet-V2S provide better accuracy as demonstrated in this study with an analysis using metrics and feature visualization. Both models attain a maximum test accuracy of ${91.78\%}$ and the minimum test loss of, with ResNet-50 at 0.24 and EfficientNet-V2S at 0.23. In addition, these models demonstrate superior precision, specificity, sensitivity, and F1-score, all around 92. Further, feature maps and class activation maps for ResNet-50 and EfficientNet-V2S show distinct and concentrated zones of activation, indicating accurate interpretation. Thus, in this comparative study, ResNet-50 and EfficientNet-V2S are choosen as the best models for this classification task owing to their high accuracy and reliable interpretability via visual activation patterns. Consequently, the mobile app ANSAN has been integrated with the ideal clinical

关键词： Glaucoma Deep learning Visualization Accuracy Sensitivity Explainable AI Computational modeling Transfer learning Decision making Visual impairment

Finger Vein Biometric Identification Using Discretization Method 2

学校读者我要写书评

暂无评论

Finger Vein Biometric Identification Using Discretization Me...

2nd International Conference on Emerging Electrical Energy, Electronics and computing Technologies 2020, ICE4CT 2020

作者： Yahaya, Y.H. Leng, W.Y. Shamsuddin, S.M. Computer Science Department Faculty of Science and Defence Technology Universiti Pertahanan Nasional Malaysia Kuala Lumpur53100 Malaysia School of Computing Faculty of Engineering Universiti Teknologi Malaysia Skudai JB83100 Malaysia UTM Big Data Centre Universiti Teknologi Malaysia Skudai JB83100 Malaysia

Over the past years, finger vein identification has gaining increasing attention in biometrics. It has many advantages as compared to other biometrics such as living-body identification, difficult to counterfeit because it resides underneath the finger skin and noninvasiveness. Finger vein feature extraction plays an important role in finger vein identification. The performance of finger vein identification is highly depending on the meaningful extracted features from feature extraction process. However, most of the works focus on how to extract the individual features and not presenting the individual characteristic of finger vein patterns with systematic representation. This paper proposed an improved scheme of finger vein feature extraction method by adopting discretization method. The extracted features will be represented systematically way in order to make classification task easier and increase the identification accuracy rate. The experimental result shows that the accuracy rate of identification of the proposed framework using Discretization is above 98.0%. © Published under licence by IOP Publishing Ltd.

关键词： Biometrics

Optimized Predictive Storage Resource Allocation Through Neural Forecasting for Cloud Environments

学校读者我要写书评

暂无评论

Optimized Predictive Storage Resource Allocation Through Neu...

Innovative computing, Intelligent Communication and Smart Electrical Systems (ICSES), International Conference on

作者： Sridevi Ponmalar. P Adithya Pothan Raj V S D Sundarsingh Jebaseelan Gayatri Phade V. Srithar Harikrishnan. R Department of Computational Intelligence School of computing Kattankulathur Campus SRM Institute of Science and Technology Chennai Dept. of CSE Saveetha School of Engineering Saveetha Institute of Medical and Technical Sciences Chennai India Dept of EEE Sathyabama Institute of Science and Technology Chennai 119 Sandip Institute of Technology & Research Centre Nashik Department of Computer Applications PeriyarManiammai Institute of Science & Technology Thanjavur Symbiosis Institute of Technology Pune Campus Symbiosis International Deemed University Pune

ISBN: (数字)9798331543624

ISBN: (纸本)9798331543631

The demand for efficient and scalable management of storage resources is growing enormously with the rapid growth in data-intensive applications within the cloud environment. Optimized PSRANF is the state-of-the-art approach towards dynamic storage resource allocation, based on the forecast of future demands in storage. PSRANF utilizes state-of-the-art neural forecasting with Long Short-Term Memory (LSTM) networks and Transformer designs to analyze the pattern of historical storage usage for highly accurate future need predictions. By proactively adjusting storage capacity to these predictions, PSRANF strikes an optimal balance between performance and cost by reserving just-in-time resources. It ensures that there is no risk of over-provisioning, which would be too costly. In the same breath, the risk of under-provisioning, which leads to performance bottlenecks, is avoided. What is more, the system will apply prioritization, placing the data of higher priority or frequently accessed data into faster storage tiers of a higher class of performance and less important classes of data into cost-effective storage. The adaptive feedback loop continuously refines the forecasting model, which, over time, improves accuracy and allows the system to handle changes in workload patterns. PSRANF combines neural forecasting with resource allocation in such a way that is cost-effective; thus, it is very well placed for cloud environments where the demand for storage is dynamic in nature and requires intelligent data-driven approaches. The AWS implementation of PSRANF has turned out to be quite a helpful tool in optimizing the storage resource allocation process. While it is able to predict 98.5% of the storage needs correctly by achieving a cost saving of up to 18.7 % , managing data from 97.8% to 99.2%, and resources from 95.4%, it's robust for improvements in cloud storage management and operational efficiency.

关键词： Cloud computing Costs Accuracy Storage management Predictive models Dynamic scheduling Resource management Forecasting Optimization Long short term memory

Pathway to Secure and Trustworthy ZSM for LLMs: Attacks, Defense, and Opportunities

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Khowaja, Sunder Ali Khuwaja, Parus Dev, Kapal Hamadi, Hussam Al Zeydan, Engin School of Computing Dublin City University CONNECT Centre DublinD09 V209 Ireland Institute of Business Administration University of Sindh Jamshoro Pakistan CONNECT Centre Department of Computer Science Munster Technological University Bishopstown CorkT12 P928 Ireland College of Engineering IT University of Dubai United Arab Emirates Barcelona08860 Spain

Recently, large language models (LLMs) have been gaining a lot of interest due to their adaptability and extensibility in emerging applications, including communication networks. It is anticipated that zero touch network and service management (ZSM) will be able to support mobile edge computing networks and LLMs as a service, as they aim to provide network automation and service management without requiring manual intervention. However, LLMs are vulnerable to data and model privacy issues that affect the trustworthiness of LLMs to be deployed for user-based services. In this paper, we explore the security vulnerabilities associated with fine-tuning LLMs in ZSM in particular the membership inference attack. We define the characteristics of an attack network that can perform a membership inference attack if the attacker has access to the fine-tuned model for the downstream task. We show that the membership inference attacks are effective for any downstream task, which can lead to a personal data breach when using LLM as a service. The experimental results show that the attack success rate of maximum 92% can be achieved on named entity recognition task. Based on the experimental analysis, we discuss possible defense mechanisms and present possible research directions to make the LLMs more trustworthy in the context of 6G networks. © 2024, CC BY-NC-ND.

关键词： Differential privacy

Wireless Power Transfer Technologies: State-Of-The-Art and Future Prospects

学校读者我要写书评

暂无评论

Wireless Power Transfer Technologies: State-Of-The-Art and F...

Communication, computing and Energy Efficient Technologies (I3CEET), International Conference on

作者： Vivek Justus S.Viveka M.Sumithra Purushothaman V V.Samuthira Pandi Desidi Narsimha Reddy Department of Computing Technologies School of Computing SRM Institute of Science and Technology Kattankulathur Tamil Nadu India Department of Computer Science and Engineering Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology Chennai Tamilnadu India Department of Electronics and Communication Engineering Chennai Institute of Technology Chennai Centre for Advanced Wireless Integrated Technology Chennai Institute of Technology Chennai Tamil Nadu India Data Consultant (Data Governance Data Analytics: Enterprise Performance Management AI&ML) Soniks Consulting LLC Texas USA

ISBN: (数字)9798331541583

ISBN: (纸本)9798331541590

Wireless Power Transfer (WPT) technologies represent a transformative approach to energy transmission, offering advantages over traditional wired systems such as convenience, flexibility, and reduced maintenance. This paper presents a comprehensive review of current advancements and future prospects in WPT. The review encompasses three primary technologies: electromagnetic induction, magnetic resonance, and microwave transmission, each offering unique benefits and challenges. Recent developments have significantly enhanced WPT efficiency, extended transmission distances, and enabled integration with emerging technologies like Internet of Things (IoT) devices and electric vehicle (EV) charging infrastructure. Looking forward, the paper identifies several key research directions and challenges. These include improving efficiency through advanced materials and metamaterials, standardization efforts to promote interoperability and safety, and exploration of new applications such as wireless charging in smart cities and space-based applications. The potential impact of WPT on reducing reliance on fossil fuels and enabling sustainable energy practices underscores its importance in future energy landscapes. By synthesizing current knowledge and highlighting future opportunities, this study aims to guide researchers, engineers, and policymakers in accelerating the development and adoption of WPT technologies, paving the way for a moreefficient and interconnected energy ecosystem.

关键词： Costs Reviews Smart cities Magnetic resonance Wireless power transfer Standardization Telecommunication computing Smart grids Internet of Things Tuning

GPT (Generative Pre-trained Transformer) - A Comprehensive Review on Enabling Technologies, Potential Applications, Emerging Challenges, and Future Directions

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Yenduri, Gokul Ramalingam, M. Selvi, Chemmalar G. Supriya, Y. Srivastava, Gautam Maddikunta, Praveen Kumar Reddy Raj, Deepti G. Jhaveri, Rutvij H. Prabadevi, B. Wang, Weizheng Vasilakos, Athanasios V. Gadekallu, Thippa Reddy The School of Information Technology and Engineering Vellore Institute of Technology Tamil Nadu Vellore632014 India The Dept. of Math and Computer Science Brandon University Canada The Research Centre for Interneural Computing China Medical University Taichung Taiwan Dept. of Computer Science and Math Lebanese American University Beirut Lebanon The Department of Computer Science and Engineering School of Technology Pandit Deendayal Energy University India The Department of Computer Science City University of Hong Kong Hong Kong Grimstad Norway Lovely Professional University Phagwara India Department of Electrical and Computer Engineering Lebanese American University Byblos Lebanon Jiaxing University Jiaxing314001 China Zhongda Group 314312 China

The Generative Pre-trained Transformer (GPT) represent a notable breakthrough in the domain of natural language processing, which is propelling us toward the development of machines that can understand and communicate using language in a manner that closely resembles that of humans. GPT is based on the transformer architecture, a deep neural network designed for natural language processing tasks. Due to their impressive performance on natural language processing tasks and ability to effectively converse, GPT have gained significant popularity among researchers and industrial communities, making them one of the most widely used and effective models in natural language processing and related fields, which motivated to conduct this review. This review provides a detailed overview of the GPT, including its architecture, working process, training procedures, enabling technologies, and its impact on various applications. In this review, we also explored the potential challenges and limitations of a GPT. Furthermore, we discuss potential solutions and future directions. Overall, this paper aims to provide a comprehensive understanding of GPT, enabling technologies, their impact on various applications, emerging challenges, and potential solutions. © 2023, CC BY.

关键词： Deep neural networks

Meta-learning Enhanced Next POI Recommendation by Leveraging Check-ins from Auxiliary Cities

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Wang, Jinze Zhang, Lu Sun, Zhu Ong, Yew-Soon Macquarie University Balaclava Rd Macquarie ParkNSW2109 Australia Chengdu University of Information Technology Chengdu China Centre for Frontier AI Research A*STAR Singapore Institute of High Performance Computing A*STAR Singapore School of Computer Science and Engineering Nanyang Technological University Singapore

Most existing point-of-interest (POI) recommenders aim to capture user preference by employing city-level user historical check-ins, thus facilitating users' exploration of the city. However, the scarcity of city-level user check-ins brings a significant challenge to user preference learning. Although prior studies attempt to mitigate this challenge by exploiting various context information, e.g., spatio-temporal information, they ignore to transfer the knowledge (i.e., common behavioral pattern) from other relevant cities (i.e., auxiliary cities). In this paper, we investigate the effect of knowledge distilled from auxiliary cities and thus propose a novel Meta-learning Enhanced next POI Recommendation framework (MERec). The MERec leverages the correlation of check-in behaviors among various cities into the meta-learning paradigm to help infer user preference in the target city, by holding the principle of "paying more attention to more correlated knowledge". Particularly, a city-level correlation strategy is devised to attentively capture common patterns among cities, so as to transfer more relevant knowledge from more correlated cities. Extensive experiments verify the superiority of the proposed MERec against state-of-the-art algorithms. Copyright © 2023, The Authors. All rights reserved.

关键词： Recommender systems