检索结果-内蒙古大学图书馆

2024 International Conference on system, Computation, Automation and Networking, ICSCAN 2024

作者： Mathivanan, P. Mahalakshmi, K. Anna University Faculty of Information and Communication Engineering Chennai India Kit - Kalaignarkarunanidhi Institute of Technology Department of Computer Science and Engineering Coimbatore India

ISBN: (纸本)9798331510022

A biometric identifier, in contrast to other identifiers used for authentication, is a quantitative evaluation of an individual's physical attributes that is successfully used to confirm or validate the identification. Because of its remarkable and consistent texture variation, iris recognition is thought to be the most dependable biometric recognition. Concerns about security, authentication, and identification are growing in importance across all domains as technology advances daily. High-security applications use these unique patterns for iris recognition. To increase the security of biometric applications without jeopardizing individual privacy, this research suggests a novel FL technique for privacy-preserving iris detection. Traditional biometric systems rely on centralised data storage and processing, which threatens data security and misuse. Federated Learning: decentralized model training that allows multiple federated clients (like mobile devices or edge servers) to collaboratively train a global model without sharing their raw iris data. We explore how this architecture can be applied to iris detection systems, maintaining accuracy while mitigating privacy risks. Our method ensures that iris features are learned locally on individual devices, and only encrypted model updates are shared with the central server. To strengthen privacy, we incorporate differential privacy and secure aggregation techniques to prevent data leakage during model training. Experimental results demonstrate the effectiveness of the proposed framework in maintaining high detection accuracy, low communication overhead, and robust privacy guarantees. © 2024 IEEE.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Enhancing IoT Security: A Study on Hybrid Intrusion Detection Methods 3

Enhancing IoT Security: A Study on Hybrid Intrusion Detectio...

引用

3rd IEEE World Conference on Applied Intelligence and Computing, AIC 2024

作者： Ali, Rabea M. Ram Baheti, Mansi Dr Babasaheb Ambedkar Marathwada University Department of Computer Science and Information Technology Maharashtra Aurangabad India

ISBN: (纸本)9798350384598

This electronic document is a 'live' template and already defines the components of your paper [title, text, heads, etc.] in its style sheet. *CRITICAL: Do Not Use Symbols, Special Characters, Footnotes, or Math in Paper Title or Abstract. (Abstract) The rapid proliferation of Internet of Things (IoT) devices has led to increased vulnerability to cyber intrusion threats like malware and denial-of-service attacks, which can severely impact system availability, reliability, and trustworthiness. Existing intrusion detection systems often rely on single environments for training models, limiting their real-world versatility against evolving attacks spanning diverse systems. To address these challenges, this study proposes an integrated deep learning approach that combines heterogeneous datasets using representation learning based on stacked convolutional autoencoders and distribution alignment. This approach aims to improve model generalization and enable the detection of threats across various IoT environments. The proposed method employs hybrid convolutional neural networks (CNN) and long short-term memory (LSTM) networks to perform robust binary and multi-class classification. By analyzing spatial context and sequential traffic dynamics, the model can distinguish normal functionality from different attack types. Extensive evaluation on the NSL-KDD and SDN-5-IoT intrusion detection benchmarks demonstrates over 99% accuracy in detecting both known and zero-day attacks. Comparisons with standalone deep architectures indicate consistent performance gains from the synergistic fusion of spatial and temporal modeling. The proposed approach advances adaptive threat intelligence for securing real-world IoT ecosystems against evolving cyber threats. The integration of heterogeneous datasets and the combination of CNN and LSTM architectures enable the model to learn more generalized representations of network traffic patterns, enhancing its ability to detect intrusions across d

关键词： Network intrusion

来源：评论

学校读者我要写书评

暂无评论

Machine Learning Techniques for Diabetes Prediction: A Comparative Analysis

Machine Learning Techniques for Diabetes Prediction: A Compa...

引用

2023 International Conference on Sustainable Computing and Smart systems, ICSCSS 2023

作者： Mythili, J. Surendhar, T. Suryaprakash, P. Suresh Kumar, K. Faculty of Computer Science and Engineering K S Rangasamy College of Technology Department of Computer Science and Engineering Tiruchengode India Computer Science and Engineering K S Rangasamy College of Technology Department of Computer Science and Engineering Tiruchengode India

ISBN: (纸本)9798350333602

As the prevalence of diabetes mellitus rises, more families are being impacted. Most diabetics don't know much about their health situation or the risks they face before getting a diagnosis. A unique data mining-based method was suggested for the prediction of type 2 diabetes mellitus (T2DM) in this study. The accuracy and robustness of the prediction model are the primary priorities, ensuring flexibility so that it may be used to other datasets. The model is made up of two parts: the enhanced Logistic Regression with Multilayer Perception (LRMLP) and Naïve Bayes (NB) method with a set of preprocessing techniques. To compare the findings with those of other studies, the Diabetes Dataset for Pima Indians was used with the Environment for Knowledge Analysis toolkit. According to the findings, the model's prediction accuracy is higher than that of earlier research. Additionally, this approach guarantees that the dataset quality is enough. The system was applied to two more diabetes datasets in order to assess its effectiveness further. Both research findings show strong performance. The model is shown to be helpful for managing diabetic health in a realistic manner as a result. For experimental analysis, data from the University of California, Irvine's (UCI) machine learning repository's Pima Indian Diabetes Database (PIDD) is used. These classification algorithms' performance is evaluated using a variety of statistical metrics, including sensitivity (recall), precision, F-score, and accuracy. The measurement of correctly and erroneously classified cases is called accuracy. According to the testing findings, LR-MLP outscored other classification algorithms and provides the maximum accuracy of 96.3%. © 2023 IEEE.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

Dialectal Arabic Speech Recognition using CNN-LSTM Based on End-to-End Deep Learning 2

Dialectal Arabic Speech Recognition using CNN-LSTM Based on ...

引用

2nd International Conference on Emerging Smart Technologies and Applications, eSmarTA 2022

作者： Alsayadi, Hamzah A. Al-Hagree, Salah Alqasemi, Fahd A. Abdelhamid, Abdelaziz A. Ibb University Faculty of Sciences Computer Science Department Ibb Yemen University of Science and Technology Department of Information Technology Yemen Ain Shams University Faculty of Computer and Information Science Computer Science Department Cairo Egypt

ISBN: (数字)9781665461818

ISBN: (纸本)9781665461818

Arabic dialect is the type of Arabic language which is used for daily communication in the Arab world. Each Arab country has a unique dialect. Due to the challenges of recognizing the spoken Arabic dialects, such as the significant variations in Arabic dialects and the lack of standard structure, there are few research attempts in the literature that address this problem. The end-to-end deep learning approach provides a promising approach for improving the performance of speech recognition systems. However, with deep learning techniques, overfitting is still the main problem with little data. In this paper, we investigate the end-to-end model to improve dialectal Arabic speech recognition (DASR) based on deep learning. Data augmentation is one of the significant steps in the proposed approach that is applied to the employed dataset to increase training data. The proposed approach is based on a hybrid model that is composed of a convolutional neural network (CNN) and long short-term memory (LSTM) and is referred to as (CNN-LSTM). This model is utilized along with attention-based encoder-decoder methods for building the acoustic model and performing the decoding process. To the best of our knowledge, there is no prior research that employed CNN-LSTM and attention-based models in dialectal Arabic ASR systems. In addition, the language model is built based on recurrent neural network (RNN) and LSTM methods. The proposed approach is validated using two speech datasets, namely SASSC and MGB-3. Experimental results showed that the proposed approach achieved a word error rate (WER) of 57.02%. This result is superior to those achieved by the other approaches in the literature. Lower WER often indicates that the ASR system is more accurate. © 2022 IEEE.

关键词： Long short-term memory

来源：评论

学校读者我要写书评

暂无评论

Text Prediction Zero Probability Problem Handling with N-gram Model and Laplace Smoothing

Text Prediction Zero Probability Problem Handling with N-gra...

引用

2022 International Conference on Advancement in Electrical and Electronic Engineering, ICAEEE 2022

作者： Mimi, Raonak Jahan Masud, Md Abdul Rahman, Rifat Dina, Nusrat Sultana Patuakhali Science and Technology University Faculty of Computer Science and Engineering Bangladesh Patuakhali Science and Technology University Department of Computer Science and Information Technology Bangladesh

ISBN: (数字)9781665469449

ISBN: (纸本)9781665469449

In Natural Language Processing, text prediction represents the process of predicting the word with the highest probability through a predictive language model from a series of text corpus. The N-gram model is familiar and considered the handiest and most computationally cost-effective model for text processing. Additionally, higher N-gram models, especially the 5-gram ones, give the best text prediction. Interestingly, these better prediction results were obtained only on the training dataset. In contrast, the highest N-gram model imploded badly on the evaluation dataset. This paper proposes an approach where the N-gram model, especially the bi-gram model, and the fine tuning with Laplace Smoothing, provide the best prediction results at the evaluation stage. © 2022 IEEE.

关键词： Text processing

来源：评论

学校读者我要写书评

暂无评论

An Ensemble based Fraudulent Blockchain Account Detection system 2024

An Ensemble based Fraudulent Blockchain Account Detection Sy...

引用

16th International Conference on Contemporary Computing, IC3 2024

作者： Ralli, Rishabh Jugran, Gargi Gaurav, Mayank Goyal, Mukta Department of Computer Science & Engineering Jaypee Institute of Information Technology U.P Noida India

ISBN: (纸本)9798400709722

This paper addresses the issue of fraud in blockchain transactions, which can significantly undermine trust and financial stability within blockchain networks. By leveraging advanced ensemble learning techniques, it aims to detect and prevent fraudulent activities effectively. The analysis focuses on blockchain networks, utilizing a combination of sophisticated machine learning algorithms such as Support Vector Machine (SVM), random forests, Tabnet, XGBoost, and the deep learning model MLP. These algorithms meticulously analyze transaction data for unusual patterns indicating fraud. The ensemble approach distinguishes between legitimate and fraudulent accounts by accurately analyzing transaction patterns. Notably, the ensemble model achieved a testing accuracy of 93.6%, thereby aiming to reduce the potential financial and reputational risks associated with fraud in blockchain transactions. © 2024 Copyright held by the owner/author(s).

关键词： Fraudulent transactions

来源：评论

学校读者我要写书评

暂无评论

Banana Leaf Disease Detection using Advanced Convolutional Neural Network

Banana Leaf Disease Detection using Advanced Convolutional N...

引用

2023 International Conference on Sustainable Computing and Smart systems, ICSCSS 2023

作者： Mohanraj, E. Hariharan, R. Keerthana, S. Cavin, K.B. Faculty of Computer Science and Engineering K S Rangasamy College of Technology Department of Computer Science and Engineering Tiruchengode India Computer Science and Engineering K S Rangasamy College of Technology Department of Computer Science and Engineering Tiruchengode India

ISBN: (纸本)9798350333602

Any plant's ability to grow disease-free is crucial for both the environment and human existence. Nevertheless, various diseases, viruses, and fungi affect the plant and highly influence the yield quality and production quantity. If the disease and its origin are not appropriately diagnosed, illness control procedures might be a waste of effort and resources, causing additional plant losses. Plant disease tracking by manual process is quite challenging. It requires a considerable amount of work, expertise in plant diseases, and protracted processing durations. Traditional neural networks require a large number of parameters to train effectively, which can make them slow and computationally expensive. The previous study includes the leaf disease detection performed by CNN, R-CNN, and SVM models. Using the Faster R-CNN, which outperforms other models in terms of computation time and accuracy, this study intends to develop a banana leaf disease detection system. © 2023 IEEE.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Traffic Sign Recognition using Convolutional Neural Networks YOLOv3 Algorithm

Traffic Sign Recognition using Convolutional Neural Networks...

引用

2023 International Conference on Self Sustainable Artificial Intelligence systems, ICSSAS 2023

作者： Gnanabaskaran, A. Pranav Ram, K.S. Vaishnavi, N. Shabari Prasannaa, D. K S Rangasamy College of Technology Faculty of Computer Science and Engineering Department of Computer Science and Engineering Tiruchengode India K S Rangasamy College of Technology Computer Science and Engineering Department of Computer Science and Engineering Tiruchengode India

ISBN: (纸本)9798350300857

Nowadays, traffic sign recognition is disrupted through various external factors such as chromatic aberration, geographical separation, and brightness of lights. This eventually poses possible safety hazards during navigation in real life. To overcome this issue, a deeplearning based method is proposed in this research using YOLOv3 with DenseNet as the backbone design. This approach uses attention mechanism to accelerate the feature extraction process by minimizing the overhead due to high computations to optimize the latency problem. Secondly, it incorporates location-specific hierarchical network to increase the traffic sign detection performance in challenging circumstances such as magnitude and illumination variations while maintaining adaptability and flexibility. In addition to this Cross Stage Partial Network is fused with DenseNet to enhance the recognition accuracy and effectiveness of the proposed system. The experimental findings on the Chinese Traffic sign dataset, CCTSDB dataset, showed that the YOLOv3-DenseNet model exhibited significant detection efficiency of 97.6% than the other deep learning models considered for performance comparison. It is observed that the proposed model is capable enough to detect the traffic signs in complicated scenarios as well as in places where the traffic signs are placed in smaller dimensions. © 2023 IEEE.

关键词： Traffic signs

来源：评论

学校读者我要写书评

暂无评论

Liver Tumour Detection through Artificial Intelligence: A Comprehensive Study 8

Liver Tumour Detection through Artificial Intelligence: A Co...

引用

8th International Conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud), I-SMAC 2024

作者： Varalakshmi, V. Hemamalini, U. Vistas Computer Science Chennai Pallavaram India Vistas Department of Information Technology Chennai Pallavaram India

ISBN: (纸本)9798350376425

Liver tumours are a serious and life-threatening condition, which, if left untreated, can lead to fatal outcomes. Given the various challenges in early disease prediction, chronic impacts, and unbalanced diagnostic information, this study focuses on creating a comprehensive analysis of liver tumour segmentation techniques. It explores the roles of machine learning, deep learning, and neural networks in diversifying early detection frameworks. The study includes an examination of feature extraction techniques used in existing frameworks, emphasizing high-quality annotations. It also addresses the limitations of current datasets and the impact of real-time databases. Detailed insights into liver tumour segmentation methods are provided, highlighting the potential of artificial intelligence frameworks for future advancements. The primary goal of this system is to develop an efficient and robust framework for detecting, segmenting, and analysing liver tumours, with a key emphasis on early detection. © 2024 IEEE.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

A Hybrid Neural Network and Box-Jenkins Models for Time Series Forecasting

引用

computers, Materials & Continua 2022年第3期70卷 4829-4845页

作者： Mohammad Hadwan Basheer M.Al-Maqaleh Fuad N.Al-Badani Rehan Ullah Khan Mohammed A.Al-Hagery Department of Information Technology College of ComputerQassim UniversityBuraydahSaudi Arabia Department of Computer Science College of Applied SciencesTaiz UniversityTaizYemen Intelligent Analytics Group(IAG) College of ComputerQassim UniversityBuraydahSaudi Arabia Faculty of Computer Sciences and Information Systems Thamar UniversityThamarYemen Faculty of Administration Science Ibb UniversityIbbYemen Department of Computer Science College of ComputerQassim UniversityBuraydahSaudi Arabia

Time series forecasting plays a significant role in numerous applications,including but not limited to,industrial planning,water consumption,medical domains,exchange rates and consumer price *** main problem is insufficient forecasting *** present study proposes a hybrid forecastingmethods to address this *** proposed method includes three *** first model is based on the autoregressive integrated moving average(ARIMA)statistical model;the second model is a back propagation neural network(BPNN)with adaptive slope and momentum parameters;and the thirdmodel is a hybridization between ARIMA and BPNN(ARIMA/BPNN)and artificial neural networks and ARIMA(ARIMA/ANN)to gain the benefits of linear and *** forecasting models proposed in this study are used to predict the indices of the consumer price index(CPI),and predict the expected number of cancer patients in the Ibb Province in *** standard measures used to evaluate the proposed method include(i)mean square error,(ii)mean absolute error,(iii)root mean square error,and(iv)mean absolute percentage *** on the computational results,the improvement rate of forecasting the CPI dataset was 5%,71%,and 4%for ARIMA/BPNN model,ARIMA/ANN model,and BPNN model respectively;while the result for cancer patients’dataset was 7%,200%,and 19%for ARIMA/BPNNmodel,ARIMA/ANN model,and BPNNmodel ***,it is obvious that the proposed method reduced the randomness degree,and the alterations affected the time series with data *** ARIMA/ANN model outperformed each of its components when it was applied separately in terms of increasing the accuracy of forecasting and decreasing the overall errors of forecasting.

关键词： Hybrid model forecasting non-linear data time series models cancer patients neural networks box-jenkins consumer price index

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：