检索结果-内蒙古大学图书馆

Visualizing the Full Spectrum Optimization of K-Nearest Neighbors From Data Preprocessing to Hyperparameter Tuning and K-Fold Validation for Cardiovascular Disease Prediction

Informatica (Slovenia)

引用

Informatica (Slovenia) 2025年第2期49卷 355-374页

作者： Joseph, Jeena Kartheeban, K. Department of Computer Applications Kalasalingam Academy of Research and Education Tamilnadu Krishnankoil India Department of Computer Applications Marian College Kerala Kuttikkanam India Computer Science and Engineering Kalasalingam Academy of Research and Education Tamilnadu Krishnankoil India

Cardiovascular disease (CVD) is a prominent cause of death worldwide. This alarming need requires an accurate prediction model using machine learning that can detect and help prevent or mitigate the risk. This study focuses on this issue and has come up with new dimensional capabilities to enhance the K-Nearest Neighbors (KNN) algorithm to predict cardiovascular diseases at an early stage by incorporating various techniques for data preprocessing and feature selection thereby improving the efficiency of the model. The proposed model identifies the most relevant features using Principal Component Analysis. The main innovation revolves around fine tuning the hyperparameter of K-Nearest Neighbors, specifically the choice of neighbors (K), using a data driven approach to ensure accuracy across different datasets. The performance of the optimized K-Nearest Neighbors algorithm is evaluated using the Framingham heart disease dataset. This model achieved an impressive prediction accuracy of 92.46% and outperformed methods that solely rely on traditional K-Nearest Neighbors. As machine learning techniques plays an important role in the development of prediction models for early detection and prevention of cardiovascular disease, this model can be considered as a valuable tool for healthcare professionals and researchers. The core contribution of this study lies in offering a comprehensive optimization of the traditional K-Nearest Neighbors (KNN) algorithm. This includes robust data preprocessing using the Hampel filter for outlier removal, feature selection through Principal Component Analysis (PCA), and performance enhancement using grid search for hyperparameter tuning combined with 10-fold cross-validation. Unlike prior studies that apply KNN with minimal adjustments, this research emphasizes the importance of an end-to-end machine learning pipeline. This holistic refinement significantly improves the predictive performance and reliability of KNN for cardiovascular diseas

关键词： Nearest neighbor search

来源：评论

学校读者我要写书评

暂无评论

A transformer based real-time photo captioning framework for visually impaired people with visual attention

引用

Multimedia Tools and Applications 2024年第41期83卷 88859-88878页

作者： Muhammed Kunju, Abubeker Kiliyanal Baskar, S. Zafar, Sherin A R, Bushara S, Rinesh A, Shafeena Karim Kerala Kanjirapally India Faculty of Engineering Department of Electronics and Communication Engineering Karpagam Academy of Higher Education Coimbatore India Department of Computer and Electronics Engineering SEST- Jamia Hamdard New Delhi India Department of Electronics and Communication Engineering KMEA College of Engineering Kerala Cochin India Department of Computer Science and Engineering V.S.B College of Engineering Technical Campus Coimbatore India

In recent years, transformer-based photo captioning frameworks plays a crucial role in improving individuals’ overall well-being, self-reliance, and inclusivity by giving them access to visual content via written and voiced explanations. This research investigates a two-layer transformer architecture to capture extensive relationships in visual data effectively. Using a two-layer transformer design, visual attention mechanism enables the model to autonomously identify relevant regions within a picture while generating a particular word within the anticipated caption. To get the features from the images, a pre-trained Inception V3 model is used, and the feature maps are extracted from the final convolutional block. The dataset used in this research was Flick8k, Flick30k, among bespoke datasets. The model was trained on a TPU platform and implemented on a Raspberry Pi 4B graphics processing unit. To assess the efficacy of the two-layer transformer model for photo captioning, a series of experiments were undertaken using benchmark datasets and with state-of-the-art models. The quality of the generated captions was evaluated using quantitative criteria such as BLEU, BLEUR, ROUGE and F1 score. Furthermore, a series of qualitative assessments were conducted, employing human evaluators, to measure the overall coherence, and fluency of the captions. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2024.

关键词： computer vision

来源：评论

学校读者我要写书评

暂无评论

Predictive HR: Ensemble & Deep Learning Methods for Strategic Employee Retention 3

Predictive HR: Ensemble & Deep Learning Methods for Strategi...

引用

3rd International Conference on Electrical, Electronics, Information and Communication Technologies, ICEEICT 2024

作者： Begum, Sheik Ariffa Nammalwar, Subbulakshmi Vijayalakshmi, Jeyapal Vallabha, Manju Kalasalingam Academy of Research and Education Computer Science and Engineering Krishnankoil India Vaigai College of Engineering Computer Science and Engineering Madurai India

ISBN: (纸本)9798350369083

In contemporary organizational landscapes, the strategic retention of employees stands as a paramount concern for sustained success. To address this challenge, this research endeavors to harness the power of predictive analytics, employing a multifaceted approach integrating ensemble learning and deep learning methodologies. The preprocessing phase employs Min-Max Scaler to normalize the feature space, enhancing the efficacy of subsequent algorithms. Additionally, Synthetic Minority Over-sampling Technique (SMOTE) is employed to mitigate class imbalances, ensuring robustness in predictive modeling. Ensemble methods such as CatBoost, AdaBoost, Light GBM, and Gradient Boosting Machine (GBM) are leveraged to capture complex interactions and nonlinear relationships within employee data. Furthermore, Long Short-Term Memory (LSTM), a variant of recurrent neural networks, is utilized to uncover temporal dependencies and intricate patterns inherent in longitudinal employee records Ultimately, this endeavor contributes to the advancement of HR analytics, empowering organizations to preemptively address attrition risks and foster a resilient and engaged workforce. © 2024 IEEE.

关键词： Predictive analytics

来源：评论

学校读者我要写书评

暂无评论

Optimization of LEACH Protocol for WSNs in Terms of Energy Efficient and Network Lifetime

引用

Journal of Cyber Security and Mobility 2023年第3期12卷 275-295页

作者： Salman, Fadhil Mohammed Mohammed, Ahssan Ahmmed Mutar, Ahmed Fakhir Ministry of Education Babylon Education Directorate Iraq Department of Computer Technology Engineering Al-Mustaqbal University College Iraq

Wireless Sensor Network (WSN) is a group of small, intelligent sensors with limited resources. WSN has limited energy restrictions, so, the network lifetime is the major challenge that directly affect the efficiency of the network. This work presents an energy-saving clustering hierarchical algorithm for WSNs;it is an improvement of Low-Energy adaptive Clustering Hierarchy (LEACH) algorithm. The aim of this algorithm is to minimize power consumption by the appropriate election of new cluster heads in every data transfer round and avoid network collisions. This goal achieved by using an efficient function to select the best cluster heads nodes in each round, which takes into account the current energy in the sensors. The proposed algorithm improves the cluster formation process by relying on the shorter distance to the base station. The Time Division Multiple Access (TDMA) mechanism also utilized to schedule the transmission of data packets to cluster heads nodes and to avoid data packet collisions at the base station. Experiments conducted in MATLAB R (2020a) software showed that the suggested algorithm extended the network lifetime by 14.5%, and improved the network throughput by 16.8% compared to the LEACH algorithm. That means, the proposed energy-saving clustering hierarchy algorithm has improved the performance of the LEACH algorithm in term of enhancing network lifetime and increasing network throughput. © 2023 River Publishers.

关键词： MATLAB

来源：评论

学校读者我要写书评

暂无评论

Deep Learning Techniques for a Comparative Study of Crop Disease Detection 5th

Deep Learning Techniques for a Comparative Study of Crop Dis...

引用

5th International Conference on Advances in Information Communication Technology and Computing, AICTC 2024

作者： Soumya Prasad, S. Sampath Kumar, L. Mallem, Sai Nirupam Gutta, Hemanth Ahmed, Rafeeq Computer Science and Engineering Koneru Lakshmaiah Education Foundation Vijayawada India Computer Science and Engineering Government Engineering College West Champaran Bettiah India

ISBN: (纸本)9789819761050

Agriculture industry can be significantly affected by pests. Deep learning techniques are very useful for detecting various diseases in crops. A dataset consisting of 1530 images was divided into the training set, testing set, and validation set. It consists of labels such as "Healthy", "Powdery," and "Rusty." These labels reflect the conditions of a plant and are used to train the model. The trained models using Deep Convolutional Neural Networks successfully find the relevant crop diseases with a 95% accuracy rate. The early detection of diseases is useful to curb the diseases in the budding stage to avoid the rampant onset of plant diseases which prove fatal to the agricultural field. By implementing this strategy, crop loss can be reduced significantly. In this paper, a detailed explanation of various CNN architectures such as Xception, VGG16, Resnet50, DenseNet, efficientNet, and InceptionV3 is outlined. Each model is trained with training data to predict a plant’s disease. The model’s performance is highly dependent on extracting features from the data. The result analysis and visualizations of obtained results are presented in the analysis section of the paper. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

ETFIDS: An Entropy-Driven, Time-Frequency Analysis Framework for In-Vehicle CAN Signal Intrusion Detection

引用

IEEE Internet of Things Journal 2025年第12期12卷 21507-21522页

作者： Liu, Wanning Qin, Guihe Liang, Yanhua Song, Jiaru Liu, Qingxin Zhou, Xue Jilin University Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education College of Computer Science and Technology Changchun130012 China

In recent years, cyberattacks against automobiles have exposed significant security threats to in-vehicle networks. The vulnerability of communication signals to malicious interference and manipulation can lead to serious system failures or abnormal behavior. The existing in-vehicle network intrusion detection methods do not fully exploit the time-frequency characteristics of controller area network (CAN) signals. This limitation reduces their effectiveness in capturing subtle changes and signal complexity. Based on the above motivation, from the perspective of signal perception, we propose an entropy-driven, time-frequency analysis framework for in-vehicle network intrusion detection. The framework integrates a signal sampler, a frequency-domain detector, and a time-domain detector. The signal sampler, as the system’s front-end module, extracts real-time physical signal data streams from CAN messages. The frequency-domain detector identifies frequency components, detecting high-frequency disturbances and cyclic variations. Meanwhile, the time-domain detector captures instantaneous changes and sudden anomalies. It analyzes signal complexity and anomalies through both stream and block detection. Experimental results demonstrate that the proposed method performs well under various attack scenarios, offering superior detection and real-time performance. It effectively senses multiple signal anomalies, providing a robust intrusion detection solution for modern in-vehicle networks. © 2014 IEEE.

关键词： Network intrusion

来源：评论

学校读者我要写书评

暂无评论

An Influence Maximization Algorithm Based on Improved K-Shell in Temporal Social Networks

引用

computers, Materials & Continua 2023年第5期75卷 3111-3131页

作者： Wenlong Zhu Yu Miao Shuangshuang Yang Zuozheng Lian Lianhe Cui College of Computer and Control Engineering of Qiqihar University Qiqihar161006China College of Teacher Education of Qiqihar University Qiqihar161006China

Influence maximization of temporal social networks(IMT)is a problem that aims to find the most influential set of nodes in the temporal network so that their information can be the most widely *** solve the IMT problem,we propose an influence maximization algorithm based on an improved K-shell method,namely improved K-shell in temporal social networks(KT).The algorithm takes into account the global and local structures of temporal social ***,to obtain the kernel value Ks of each node,in the global scope,it layers the network according to the temporal characteristic of nodes by improving the K-shell ***,in the local scope,the calculation method of comprehensive degree is proposed to weigh the influence of ***,the node with the highest comprehensive degree in each core layer is selected as the ***,the seed selection strategy of KT can easily lose some influential ***,by optimizing the seed selection strategy,this paper proposes an efficient heuristic algorithm called improved K-shell in temporal social networks for influence maximization(KTIM).According to the hierarchical distribution of cores,the algorithm adds nodes near the central core to the candidate seed *** then searches for seeds in the candidate seed set according to the comprehensive *** showthatKTIMis close to the best performing improved method for influence maximization of temporal graph(IMIT)algorithm in terms of effectiveness,but runs at least an order of magnitude faster than ***,considering the effectiveness and efficiency simultaneously in temporal social networks,the KTIM algorithm works better than other baseline algorithms.

关键词： Temporal social network influence maximization improved K-shell comprehensive degree

来源：评论

学校读者我要写书评

暂无评论

Fake and propaganda images detection using automated adaptive gaining sharing knowledge algorithm with DenseNet121

引用

Journal of Ambient Intelligence and Humanized Computing 2024年第9期15卷 3519-3531页

作者： Muthukumar, A. Raj, M. Thanga Ramalakshmi, R. Meena, A. Kaleeswari, P. Department of Electronics and Communication Engineering Kalasalingam Academy of Research and Education Tamil Nadu Krishnankoil India Department of Computer Science and Engineering Kalasalingam Academy of Research and Education Tamil Nadu Krishnankoil India Department of Electronics and Communication Engineering KLN College of Engineering Tamil Nadu Pottapalayam India

An additional tool for swaying public opinion on social media is to present recent developments in the creation of natural language. The term "Deep fake" originates from deep learning technology, which effortlessly and seamlessly steers someone toward digital media. Artificial Intelligence (AI) techniques are a crucial component of deep fakes. The generative powers of generative capabilities greatly reinforce the advancements in language modeling for content generation. Due to low-cost computing infrastructure, sophisticated tools, and readily available content on social media, deep fakes propagate misinformation and hoaxes. These technologies make it simple to produce misinformation that spreads fear and confusion to everyone. As such, distinguishing between authentic and fraudulent content can be challenging. This study presents a novel automated approach for the identification of deep fakes, based on Adaptive Gaining Sharing Knowledge (AGSK) and using DenseNet121 architecture. During pre-processing, the image’s sensitive data variance or noise is eliminated. Following that, CapsuleNet is used to extract the feature vectors. The deep fake is identified from the resulting feature vectors by an AGSK with DenseNet121 architecture, together with the hyper-parameter that has been optimized using the AGSK model. Propaganda and defamation pose less of a concern, and the results of the suggested deepfake image recognition model show how reliable and successful the model is. The accuracy of detection is 98% higher than other cutting-edge models. © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2024.

关键词： Digital storage

来源：评论

学校读者我要写书评

暂无评论

SEEK Mobility Adaptive Protocol Destination Seeker Media Access Control Protocol for Mobile WSNs

Iraqi Journal for Computer Science and Mathematics

引用

Iraqi Journal for computer Science and Mathematics 2023年第1期4卷 130-145页

作者： Shaker, Atheel Sabih Youssif, Omar F. Aljanabi, Mohammad Abbood, Zainab Ali Mahdi, Mahdi S. Computer Engineering Techniques Baghdad College of Economic SciencesUniversity Baghdad Iraq Computer Engineering Techniques Ibn Khaldun University College Baghdad Iraq Department of computer science Alsalam university college Baghdad Iraq Department of computer Aliraqia university college of education Baghdad Iraq Computer Technology Engineering dep Al-Esraa University College Baghdad Iraq Department of computer technology engineering Al-Esraa University college Baghdad Baghdad Iraq

The mobile wireless sensor network (WSN) is an emerging field as it widens the scope of applications where sensor networks can be applied. Generally, mobility in wireless communications degrades how well each pair of nodes communicate with one another. In WSN, the effect is higher because nodes have limited communication and computational capabilities. Those limitations create a challenging environment for the operation of sensor node communications. This paper proposes an energy-efficient media access control (MAC) protocol with mobility adaptive throughput based on a carrier-sense multiple access with collision avoidance MAC mechanism called the SEEK-mobility adaptive protocol (SEEK-MADP). SEEK-MADP uses a unique control packet operation to transfer data packets through as many nodes as possible in one duty cycle. The control packets, SYNC and RTS, are merged into one packet (SEEK). SEEK is then transferred to the downstream nodes to establish a connection between a stream of nodes in a duty cycle. This process minimizes the energy consumed by the handshaking process between the connected nodes. To increase throughput, the data period is adaptive to minimize/maximize the data packets according to the movement speed of the sender/receiver nodes. The proposed algorithm is assessed via extensive simulations in mobile scenarios using Network Simulator version 2. The final results show that the SEEK-MADP outperforms MAC protocols, sensor MAC (S-MAC), and SMAC with adaptive listening, which have shown good performance in mobile scenarios. The performance of the proposed algorithm is better than that of IEEE 802.15.4 standard MAC mechanism at mobile scenarios. © 2023 Authors. All rights reserved.

关键词： Medium access control

来源：评论

学校读者我要写书评

暂无评论

Monitoring the security of audio biomedical signals communications in wearable IoT healthcare

引用

Digital Communications and Networks 2023年第2期9卷 393-399页

作者： Saeid Yazdanpanah Saman Shojae Chaeikar Alireza Jolfaei Department of Computer Engineering Khorramabad BranchIslamic Azad UniversityKhorramabadIran Australian Institute of Higher Education SydneyAustralia College of Science and Engineering Flinders UniversityAdelaideAustralia

Smart home devices are vulnerable to a variety of *** matter gets more complicated when a number of devices collaborate to launch a colluding attack(e.g.,Distributed-Denial-of-Service(DDoS))in a network(e.g.,Smart home).To handle these attacks,most studies have hitherto proposed authentication protocols that cannot necessarily be implemented in devices,especially during Device-to-Device(D2D)*** into the potential of Ethereum blockchain and smart contracts,this work proposes a lightweight authentication mechanism that enables safe D2D interactions in a smart *** Ethereum blockchain enables the implementation of a decentralized prototype as well as a peer-to-peer distributed ledger *** work also uses a single server queuing system model and the authentication mechanism to curtail DDoS attacks by controlling the number of service requests in the *** simulation was conducted twenty times,each with varying number of devices chosen at random(ranging from 1 to 30).Each requester device sends an arbitrary request with a unique resource requirement at a *** is done to measure the system's consistency across a variety of device *** experimental results show that the proposed protocol not only prevents colluding attacks,but also outperforms the benchmark protocols in terms of computational cost,message processing,and response times.

关键词： DDoS attack Device-to-device Smart homes Blockchain Authentication Ethereum smart contract

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：