检索结果-内蒙古大学图书馆

IAENG International Journal of computer Science 2021年第2期48卷 1-13页

作者： Sun, Wei-Zhong Guo, Fu-Jun Wang, Jie-Sheng Chen, Lin Wei, Dong Du, Xin-Feng A lecturer of School of Computer Science and Software Engineering University of Science and Technology Liaoning Anshan114051 China A postgraduate student of School of Electronic and Information Engineering University of Science and Technology Liaoning Anshan114051 China A professor of School of Electronic and Information Engineering University of Science and Technology Liaoning Anshan114051 China An associate professor of School of Electronic and Information Engineering University of Science and Technology Liaoning Anshan114051 China

A colony is a group of microorganisms produced by the growth and reproduction of a single microorganism species. A method of colony image feature extraction, essential dimension estimation and dimension reduction based on digital image processing technology was proposed. Firstly, based on the HSI image color moments (first, second and third order moments), nine color characteristics of the colony images were extracted. Based on gray-level co-occurrence matrix (GLCM), twenty texture features of colony images were obtained. Then, three essential dimension estimation methods, namely, the correlation dimension estimator, the maximum likelihood estimator and the packing numbers estimator, are used to estimate the dimension of the inner low-dimensional structure of the high-dimensional data of the colony images. The results of different dimension estimators were combined with six different dimension reduction techniques (PCA, LDA, MDS, ISOMAP, SNE and NCA) respectively to form 18 different data dimension reduction methods. Based on the distance criterion function, an optimal dimension reduction method for the high-dimensional feature data of the colony images corresponding to the maximum distance criterion function was obtained. © 2021. All Rights Reserved.

关键词： Feature extraction

来源：评论

学校读者我要写书评

暂无评论

GAN-CNN Ensemble: A Robust Deepfake Detection Model of Social Media Images Using Minimized Catastrophic Forgetting and Generative Replay Technique

引用

Procedia computer Science 2024年 235卷 948-960页

作者： Preeti Sharma Manoj Kumar Hitesh Kumar Sharma Research Scholar School of Computer Science University of Petroleum and Energy Studies (UPES) Dehradun 248007 India Associate Professor Faculty of Engineering and Information Sciences University of Wollongong in Dubai UAE MEU Research Unit Faculty of Information Technology Middle East University Amman 11831 Jordan

Deep-fake photographs are difficult to discern from real ones, especially when utilized in social media platforms. Anyone can willfully create disinformation about public personalities, politicians, and celebrities using these deep fake photographs. So, it is an important need of society to work for an effective model for its detection. The models for deep fake detection commonly use CNN-based detectors. These detectors experience a drop in performance when used for transfer learning or continual learning techniques. A significant limitation in this process is CNN's catastrophe forgetting defect. For the solution of this problem, a Generative replay technique in the form of a GAN-CNN model is implemented that works to minimize this catastrophe forgetting issue that further helps for better detection. It involves generating and storing samples from previous tasks and then replaying them during the training of new tasks which makes CNN more robust to identify deep fakes. The GAN model used in this work is traditional DCGAN improved with necessary adjustments to achieve training stability. It is observed that the model attained a good accuracy of 98.67%(training),70.08% (testing) and minimum loss with a value of 0. 0337 for 100 epochs. Also, it acquired good precision values of 68% and 72%, Recall values are 74% and 66%, and F1 scores of 71% and 69% for classes 0 and 1 respectively. The model outcome is found stable and reliable in deep fake detection under dynamic training conditions. Optimum values of evaluation parameters ensure the model's capacity to learn new tasks preserving the existing task-learning knowledge.

关键词： Deep Learning CNN GAN Catastrophe forgetting CNN continual learning Lifelong learning GAN-CNN deep fake detector

来源：评论

学校读者我要写书评

暂无评论

Deep Reinforcement Learning for Energy Efficiency Optimization using Autonomous Waste Management in Smart Cities

Deep Reinforcement Learning for Energy Efficiency Optimizati...

引用

Trends in Material Science and Inventive Materials (ICTMIM), International Conference on

作者： T. Kavitha Koushik Reddy Chaganti Sivananda Lahari Reddy Elicherla M. Ravi Kumar D. Chaithanya K Manikanta Assistant Professor Department of CSE Srinivasa Ramanujan Institute of Technology Research Scholar Dept. of Computer and Information Systems University of the Cumberlands Associate Professor Department of CSE Dayananda Sagar University Assistant Professor Department of Cyber Security & IoT Malla Reddy University Assistant Professor Department of CSE CVR College of Engineering Assistant Professor Department of CSE Gates Institute of Technology

ISBN: (数字)9798331501488

ISBN: (纸本)9798331501495

Smart cities face vital waste management challenges regarding efficiency and sustainability in their development period. The established waste collection approaches that depend on pre-set schedules together with unchanging routes create problems by possessing both inefficient resource usage and elevated fuel usage as well as delayed waste pickup times. The proposed framework builds a Deep Reinforcement Learning (DRL)-based system for dynamic waste collection route optimization using Autonomous Waste Collection Vehicles (AWCVs) with Internet of Things (IoT)-based smart bins. The system uses Deep Q-Networks (DQNs) to analyze traffic conditions data alongside vehicle battery level data with smart bin information from the system so vehicles carry out efficient waste collection operations using minimal energy. The newly proposed algorithm provides 30% distance reduction and 28% faster responses and 30% lower energy usage than baseline heuristic methods do. The conclusions show that AI-powered waste management strategies can create sustainable operations which increase both service performance and overall energy efficiency in cities. The research team plans to apply the proposed approach to actual waste collection systems while exploring methods of deploying multiple agents alongside renewable energy systems to enhance future urban waste management operations.

关键词： Waste management Schedules Renewable energy sources Smart cities Deep reinforcement learning Energy efficiency Real-time systems Vehicle dynamics Sustainable development Optimization

来源：评论

学校读者我要写书评

暂无评论

Bird Swarm Algorithm with Fuzzy Min-Max Neural Network for Financial Crisis Prediction

引用

computers, Materials & Continua 2022年第10期73卷 1541-1555页

作者： K.Pradeep Mohan Kumar S.Dhanasekaran I.S.Hephzi Punithavathi P.Duraipandy Ashit Kumar Dutta Irina V.Pustokhina Denis A.Pustokhin Department of Computing Technologies Associate ProfessorSRM Institute of Science and TechnologyKattankulathur603203India Department of Information Technology Kalasalingam Academy of Research and Education626126India Department of Computer Science and Engineering Sphoorthy Engineering CollegeHyderabadTelangana501510India Department of Electrical and Electronics Engineering J B Institute of Engineering and TechnologyHyderabadTelangana500075India Department of Computer Science and Information System College of Applied SciencesAlMaarefa UniversityRiyadh11597Kingdom of Saudi Arabia Department of Entrepreneurship and Logistics Plekhanov Russian University of Economics117997MoscowRussia Department of Logistics State University of Management109542MoscowRussia

Financial crisis prediction(FCP)models are used for predicting or forecasting the financial status of a company or financial *** is considered a challenging issue in the financial *** and machine learning(ML)models can be employed for the design of accurate FCP *** numerous works have existed in the literature,it is needed to design effective FCP models adaptable to different *** study designs a new bird swarm algorithm(BSA)with fuzzy min-max neural network(FMM-NN)model,named BSA-FMMNN for *** major intention of the BSA-FMMNN model is to determine the financial status of a firm or *** presented BSA-FMMNN model primarily undergoes minmax normalization to transform the data into uniformity ***,k-medoid clustering approach is employed for the outlier removal ***,the classification process is carried out using the FMMNN model,and the parameters involved in it are tuned by the use of *** utilization of proficient parameter selection process using BSA demonstrate the novelty of the *** experimental result analysis of the BSA-FMMNN model is validated using benchmark dataset and the comparative outcomes highlighted the supremacy of the BSA-FMMNN model over the recent approaches.

关键词： Financial crisis predictive model machine learning outlier removal clustering metaheuristics

来源：评论

学校读者我要写书评

暂无评论

Intrusion Detection in Novel WSN-Leach Dos Attack Dataset using Machine Learning based Boosting Algorithms

引用

Procedia computer Science 2023年 230卷 90-99页

作者： Dharini N Jeevaa Katiravan Sruthi Priya D M Sakthi Sneghaa V A Associate Professor Department of Computer Science Engineering(Cyber Security) R.M.K. College of Engineering and Technology Thiruvallur 601206 Professor Department of Information Technology Velammal Engineering College Surapet Chennai 600066 Software Engineer embedUR Chennai 600013 Jr. Customer Service Executive Tata Communications Transformation Services Chennai 600037

Security poses a substantial challenge in the context of wireless sensor networks (WSNs). WSNs, which are frequently employed for real-time monitoring in various Internet of Things (IoT)-based applications, are particularly vulnerable to a range of Denial of Service (DoS) attacks. These attacks represent a significant security concern and are among the most prevalent issues confronting WSNs *** our research, we have introduced a WSN DoS dataset using the Matrix Laboratory (MATLAB) tool. This dataset was developed through the utilization of the Leach Protocol and incorporates various types of DoS attacks, including Wormhole, Blackhole, Grayhole, Flooding, and Time Division Multiple Access (TDMA) attacks. In our study, we assessed the performance of several machine learning techniques using our dataset. We employed various performance evaluation metrics that gauge both the classification accuracy and the complexity of these machine learning models. These metrics included parameters such as the number of accurately classified and misclassified instances, along with the time required for constructing the model. This comprehensive evaluation aimed to demonstrate the effectiveness of our proposed *** results of our analysis revealed that the Light Gradient Boosting Machine (LightGBM), Extreme Gradient Boosting (XGBoost), and Bagging algorithms exhibited strong performance on our dataset. They achieved impressive accuracy levels of up to 99.16%, 99.16%, and 99.8%, respectively, when using 14 features. Additionally, these algorithms achieved accuracy rates of 99.16%, 98.75%, and 99.8%, respectively, with 5 features, and 95.41%, 95%, and 93.75%, respectively, when considering only 3 features such as node ID and Energy-Consumed, Attack_type. Notably, LightGBM outperformed the others in terms of speed, as it employs a leaf-wise growth strategy, consumes less memory, and requires less time for model construction. Consequently, LightGBM emerged as the most efficien

关键词： Denial of Service (DoS) attacks Intrusion Detection System (IDS) Leach Protocol WSN DoS dataset Ensemble Learning Feature Selection Accurately Classified and Misclassified instance Light Gradient Boosting Machine (LightGBM)

来源：评论

学校读者我要写书评

暂无评论

Intrusion detection in big data using hybrid feature fusion and optimization enabled deep learning based on spark architecture

引用

computers and Security 2022年 116卷

作者： M .P . , Ramkumar Reddy, P.V. Bhaskar Thirukrishna, J.T. Vidyadhari, Ch. Assistant Professor Department of Computer Science and Engineering Thiagarajar College of Engineering Madurai625015 India Professor School of Computer Science and Engineering REVA University Karnataka Bengalore India Associate Professor Department of Information Science and Engineering Dayananda Sagar Academy of Technology and Management Karnataka Bangalore560082 India Assistant Professor Department of Information Technology Gokaraju Rangaraju Institute of Engineering and Technology Bachupally Hyderabad India

With the rapid expansion of Internet services and increasing intrusion issues, conventional intrusion detection techniques cannot work well with several difficult intrusions. Though several intrusion detection methods have been introduced in the past years, developing an Intrusion Detection Systems (IDS) with prevailing intrusion detection strategy is still very much desirable. Hence, a robust and effective intrusion detection approach, named RV coefficient+Exponential Sea Lion Optimization-enabled Deep Residual Network (ExpSLO-enabled DRN) using spark is devised for the intrusion detection. Here, the unique features are selected using proposed RV coefficient-based hybrid feature fusion, which is designed by the incorporation of wrapper, class-wise information gain (CIG), and Canberra distance in slave node. With the distinctive features selected, the process of data augmentation is done using oversampling for making the data more appropriate to perform the further process in the slave node. Moreover, DRN classifier is utilized for detecting the intrusions in the master node where the DRN training is done using devised ExpSLO algorithm, which is the hybridization of Exponentially Weighted Moving average (EWMA) and Sea Lion Optimization (SLnO). Furthermore, the devised method obtained better performance by considering the evaluation metrics, such as precision, recall, and F-measure with the higher values of 0.8800, 0.8845, and 0.8822 based on without attacks using dataset-2. © 2022

关键词： Intrusion detection

来源：评论

学校读者我要写书评

暂无评论

Hand Gesture Identification and Voice Command Based Hardware Reduction

Hand Gesture Identification and Voice Command Based Hardware...

引用

IEEE Punecon

作者： Amogh Godbole Vishal Gondke Kailas K. Devadkar Dept. of Information Technology Sardar Patel Institute Of Technology (SPIT) Mumbai India Dept.of Information Technology Sardar Patel Institute Of Technology (SPIT) Mumbai India Ph.D (Technology) in Computer Engineering Associate Professor Dept. Of Information Technology Sardar Patel Institute Of Technology (SPIT) Mumbai India

ISBN: (纸本)9781665498982

The use of physical devices like mouse and keyboard to communicate to our computer has hindered the natural interface between the computer and its user. To eradicate this barrier, we have designed a virtual mouse which will act upon the various hand gestures the user provides and will help them interact with the computer in a better way. The virtual mouse control of our system will keep a track of the movements of the fingers and palm, thus providing us with a specific output based on the recognized action of our hand. The proposed model will recognize both static as well as dynamic hand gestures and the results will show that the model that has been proposed and built will intuitively interact with the computer and it can be achieved the minimum usage of any hardware requirements. Our model will also have a major advantage over the current existing models, which is, it also has the feature of voice command recognition, meaning that even if the user if not able to provide appropriate gestures using their hand due to some medical condition, they will still be able to interact with the computer via voice commands.

关键词： Medical conditions Tracking Computational modeling IEEE Sections Keyboards Speech recognition Control systems

来源：评论

学校读者我要写书评

暂无评论

A Novel Forward Filter Feature Selection Algorithm Based on Maximum Dual Interaction and Maximum Feature Relevance(MDIMFR) for Machine Learning

A Novel Forward Filter Feature Selection Algorithm Based on ...

引用

International Conference on Advanced computer Control, ICACC

作者： M A Anitha K K Sherly Assistant Professor of Computer Science College of Engineering Cherthala Kerala India Associate Professor of Information Technology Rajagiri School of Engineering & Technology Kerala India

ISBN: (纸本)9781665439206

In the last few decades, Feature selection is one of the most challenging and open problem to researchers. The rapid progress in computational techniques causes the generation and recording of data in huge size. Though there exists various feature ranking methods, the processing of data is still a challenging task due to its computational complexity. The filter method has many advantages over the wrapper method. The filter methods are classifier independent and have better computational efficiency. Here, a subset of features is selected based on a certain goal function. Most of these goal functions employs the principle of information theory. Most of the algorithms in earlier studies addressed two factors, that is, maximization of relevancy and minimization of redundancy without considering the interaction among the features. This paper developed a new forward filter feature selection algorithm based on mutual information known as Maximum Dual Interaction and Maximum Feature Relevance(MDIMFR). This method considers all the three factors: relevance, redundancy, and feature interaction. This method is experimented on three datasets and compares the performance with existing methods. The results show that MDIMFR outperforms the existing competitive feature selection methods of recent studies: mRMR, JMIM and CMIM. MDIMFR also achieves good stability in average classification accuracy for a certain number of features, say k and above. Hence, these k features can be considered as an optimal feature set.

关键词： Machine learning algorithms Redundancy Filtering algorithms Feature extraction information filters Minimization Filtering theory

来源：评论

学校读者我要写书评

暂无评论

A novel prediction of remaining useful life time of rolling bearings using convolutional neural network with bidirectional long short term memory

Journal of Engineering Research (Kuwait)

引用

Journal of Engineering Research (Kuwait) 2024年

作者： Balamurugan, R. Takale, Dattatray G. Parvez, M. Muzammil Gnanamurugan, S. Associate Professor Department of Mechanical Engineering M.Kumarasamy College of Engineering Karur India Department of Computer Engineering Vishwakarma Institute of Information Technology Maharashtra Pune India Koneru Lakshmaiah Education Foundation Andhra Pradesh Vaddeswaram India Assistant Professor/ ECE Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology (Deemed to be University) Avadi Chennai India

Detection of bearing problems increases the importance of the service life of rotating machinery. Convolutional neural networks (CNNs) are often used in current research, and databases built on deep learning (DL) models have improved capabilities in the field of defect diagnosis. We use the publicly available Case Western Reserve University (CWRU) dataset to compare the classification accuracy and gain more adaptive knowledge and insights about the proposed approach. Extensive tests and evaluations are performed on the dataset to verify the diagnostic effectiveness of the recommended method in different situations. To demonstrate the superiority of the proposed method, we compare multiple views of the same dataset with similar tasks. CNN supports degraded index sequences to reduce noise and stop temporal oscillations. A new CNN-BiLSTM model is used to capture current and historical inspection data and predict the RUL's service life and supported power levels. Regarding production, we follow health rates. The proposed method was evaluated by accelerating the bearing motion to failure, and the results demonstrated its advantages in terms of more accurate RUL prediction. According to the experimental results, the proposed center distance measurement method is a new and valuable means for intelligent bearing diagnosis. Experimental results using 48 K and 12 K CWRU datasets show that the overall accuracy of the BiLSTM method is 99.80% and 98.3%, respectively, which is better in diagnosis than some popular models. © 2024 The Authors

关键词： BiLSTM CNN CWRU dataset Rolling bearings life time

来源：评论

学校读者我要写书评

暂无评论

Unsupervised Single Valued Neutrosophic Sets Approach for Cloud Clustering

引用

Procedia computer Science 2025年 258卷 3580-3589页

作者： Haewon Byeon Pramod Kumar Ihtiram Raza Khan Faisal Yousef Alghayadh Maher Ali Rusho Mukesh Soni Inje University Medical Big Data Research Center Gimhae 50834 Republic of Korea Associate professor Ganga Institute of Technology and Management Jhajjar - Bahadurgarh Rd Kablana Bhadana Haryana 124104 Computer science department Jamia Hamdard Delhi Computer Science and Information Systems Department College of Applied Sciences AlMaarefa University Riyadh Saudi Arabia Department of Lockheed Martin Engineering Management University of Colorado Boulder Boulder Colorado Dr. D.Y. Patil Vidyapeeth Pune Dr. D.Y. Patil School of Science & Technology Tathawade Pune

This work presents an innovative approach to integrating restricted rationality and strategic manipulation into multi-attribute decision-making in product style design, resulting in more precise and objective outcomes. The suggested approach employs cloud model clustering and unsupervised Single-Valued Neutrosophic Sets (SVNS). SVNS is derived by constructing pairwise comparison ratio square matrices between baseline attributes and prospective alternatives, following the scientific clustering of decision experts. SSVNC denotes the degrees of real, false, and questionable membership. The relative weights of the standard attributes are established by analysing the evaluation outcomes of different solutions for each common trait. Cloud model clustering is employed to exclude conflicting and unreasonable experts, hence attaining precise weights for design judgements. This model incorporates multi-granularity linguistic factors. The integration of the significance of generic attributes and human assessment yields a composite score for each prospective response. The feasibility and efficacy of the proposed strategy are demonstrated through a case study in automotive design aesthetics. The results emphasise the method’s capability to safeguard against nefarious strategic manipulation and effectively address complex and confusing multi-attribute decision-making challenges in vehicle stylistic design.

关键词： Multi-attribute decision-making Bounded rationality Strategy manipulation unsupervised Single-Valued Neutrosophic Sets Cloud model clustering

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：