检索结果-内蒙古大学图书馆

Density Based Clustering Approach Combined with Fuzzy Logic for Elimination of Random Valued Impulse Noise

Journal of The Institution of Engineers (India): Series B 2025年 1-14页

作者： Basu, Sulagna Bandyopadhyay, Aritra Department of Computer Science and Engineering Supreme Knowledge Foundation Group of Institutions 1 Khan Road Mankundu West Bengal Chandannagar712139 India

The reduction of impulse noise is crucial in processing pictures since it directly impacts the patterns of noise present. This paper proposes a two-step technique, known as DCIFF (DBSCAN clustering identified fuzzy filter), to effectively eliminate impulsive noise from digital images. The suggested DCIFF methodology employs the DBSCAN clustering algorithm to identify and classify noise. Subsequently, a noise reduction technique is presented, which relies on the principles of fuzzy reasoning. The density-based clustering approach DBSCAN is utilized to identify commonalities throughout all 5 × 5 sized windows that are created by centering each of the pixels in the picture. Next, to properly reduce noise, a sized 7 × 7 window is formed with the centre of each identified disturbed pixel. The non-noisy pixel intensities are multiplied by their triangle fuzzy membership values, and then the products are added together. The total result is then divided by the sum of the fuzzy membership of all the non-noisy pixels’ intensities. The outcome is utilized to eliminate the noise from the central pixel that is affected by noise. The DCIFF is assessed using standard image evaluation criteria for various pictures. On the Lena image modified by random valued impulsive noise at 80% noise density, DCIFF provided peak-signal-to-noise ratios (PSNR) of 25.86 dB and structural similarity indexes (SSIM) of 0.8429. In contrast to prior research on the same topic, the current study findings consistently demonstrate striking outcomes. © The Institution of Engineers (India) 2025.

关键词： Fuzzy logic

来源：评论

学校读者我要写书评

暂无评论

Development of a machine learning model for predicting abnormalities of commercial airplanes

引用

Data science and Management 2024年第3期7卷 256-265页

作者： Rossi Passarella Siti Nurmaini Muhammad Naufal Rachmatullah Harumi Veny Fara Nissya Nur Hafidzoh Department of Computer Engineering Faculty of Computer ScienceUniversitas Sriwijaya30662Indonesia Intelligent System Research Group(ISysRG) Faculty of Computer ScienceUniversitas Sriwijaya30662Indonesia Department of Informatics Faculty of Computer ScienceUniversitas Sriwijaya30662Indonesia College of Engineering Studies Universiti Teknologi MARA(UiTM)Shah Alam40450Malaysia

Airplanes are a social necessity for movement of humans,goods,and *** are generally safe modes of transportation;however,incidents and accidents occasionally *** prevent aviation accidents,it is necessary to develop a machine-learning model to detect and predict commercial flights using automatic dependent surveillance–broadcast *** study combined data-quality detection,anomaly detection,and abnormality-classification-model *** research methodology involved the following stages:problem statement,data selection and labeling,prediction-model development,deployment,and *** data labeling process was based on the rules framed by the international civil aviation organization for commercial,jet-engine flights and validated by expert commercial *** results showed that the best prediction model,the quadratic-discriminant-analysis,was 93%accurate,indicating a“good fit”.Moreover,the model’s area-under-the-curve results for abnormal and normal detection were 0.97 and 0.96,respectively,thus confirming its“good fit”.

关键词： Automatic dependent surveillance-broadcast data Commercial airplanes accident Data-labeling Machine learning Prediction model

来源：评论

学校读者我要写书评

暂无评论

Geospatial Project: Landslide Prediction 1

引用

5th International Conference on Data and Information sciences, ICDIS 2023

作者： Sharma, Harsh Jindal, Harsh Sharma, Megha Sehgal, Abhinav Sharma, Abhinav Godha, Rohan Computer Science and Engineering Chandigarh University Kochi India Computer Science and Engineering Chandigarh Group of College Sahibzada Ajit Singh Nagar India

ISBN: (数字)9789819969067

ISBN: (纸本)9789819969050

This literature review examines the use of machine learning (ML) algorithms for landslide identification and provides an overview of recent studies in this field. The most used algorithms for landslide identification include support vector machine (SVM), decision trees, random forests, artificial neural networks (ANNs), and deep learning models such as convolutional neural networks (CNN). The review highlights the strengths and limitations of these approaches, such as data scarcity, imbalanced datasets, and interpretability issues, and proposes solutions to these challenges. Two specific studies in landslide prediction using ML are discussed, including a digital inventory using supervised learning to detect landslides and an optimized random forest model to evaluate landslide susceptibility with 16 conditioning factors. The review concludes by emphasizing the potential of ML techniques for landslide identification and the importance of understanding their strengths and limitations. This paper provides valuable insights for researchers and practitioners interested in applying ML algorithms for landslide identification and prediction. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd 2024.

关键词： Rain

来源：评论

学校读者我要写书评

暂无评论

Birch Clustering Approach for Detection of Random Valued Impulse Noise

引用

Journal of The Institution of Engineers (India): Series B 2024年 1-9页

作者： Basu, Sulagna Bandyopadhyay, Aritra Department of Computer Science and Engineering Supreme Knowledge Foundation Group of Institutions 1 Khan Road West Bengal Mankundu712139 India

computer vision relies on image processing for autonomous driving, surveillance, and medical imaging. Clustering, an unsupervised learning approach, is essential for picture data organization and smooth pre-processing. Photo noise has been removed using K-Means, K-Medoid, and Fuzzy C-Means clustering methods. K-means, K-Medoid, and Fuzzy C-Means may not cluster huge datasets well with limited memory or CPU. Traditional clustering methods struggle to accommodate running durations and quality as dataset quantities rise. Birch clustering, or Balanced Iterative Reducing and Clustering utilizing Hierarchies, is frequently used in image processing because of its scalability and efficiency. Hierarchies help BIRCH summarize the dataset while maintaining as much information as feasible. The smaller summary follows the larger dataset. BIRCH is often used alongside other clustering methods to compress the dataset for the next step. Birch clustering is scalable, efficient in high-dimensional spaces, and can handle enormous datasets. Birch clustering regularly builds a tree structure to arrange images into a hierarchy of sub-clusters for effective segmentation and representation. Birch clustering divides images into sections by examining pixel intensities or characteristics for image segmentation. Birch clustering identifies typical centroids inside clusters to simplify feature extraction and allow meaningful picture data displays. Its noise reduction and data distribution adaptability make it suited for many academic and industrial image processing tasks. Birch clustering's hierarchical tree structure allows for scalability, unlike k-means' centroids-based clusters. Birch clustering creates a hierarchical tree using image data sub-clusters and centroids. It handles massive datasets thanks to its efficient memory storage. It offers scalable and efficient clustering with decreased computing complexity by dividing and combining data to create a hierarchy. The detection capabilitie

关键词： Image segmentation

来源：评论

学校读者我要写书评

暂无评论

Characterizing the app recommendation relationships in the iOS app store: a complex network's perspective

引用

science China(Information sciences) 2025年第4期68卷 284-299页

作者： Gang HUANG Fuqi LIN Yun MA Haoyu WANG Qingxiang WANG Gareth TYSON Xuanzhe LIU Key Lab of High Confidence Software Technologies (Peking University) Ministry of Education School of Computer Science Peking University Institute for Artificial Intelligence Peking University School of Cyber Science and Engineering Huazhong University of Science and Technology Alibaba Group Hong Kong University of Science and Technology (Guangzhou)

Mobile apps have become widely adopted in our daily lives. To facilitate app discovery, most app markets provide recommendations for users, which may significantly impact how apps are accessed. However, little has been known about the underlying relationships and how they reflect(or affect) user behaviors. To fill this gap, we characterize the app recommendation relationships in the i OS app store from the perspective of the complex network. We collect a dataset containing over 1.3 million apps and 50 million app recommendations. This dataset enables us to construct a complex network that captures app recommendation relationships. Through this, we explore the recommendation relationships between mobile apps and how these relationships reflect or affect user behavior patterns. The insights gained from our research can be valuable for understanding typical user behaviors and identifying potential policy-violating apps.

关键词： mobile app recommendation complex network user behavior policy-violating app

来源：评论

学校读者我要写书评

暂无评论

Deep neural network for emojis-based sentiment classification

引用

International Journal of Information Technology (Singapore) 2025年 1-8页

作者： Balaji, Prasanalakshmi Alsid, Linda Elzubir Gasm Sagar, Kalpna Prakasam, Shanmugapriya Department of Computer Science College of Computer Science King Khalid University Abha Saudi Arabia Department of Computer Science KIET Group of Institutions Delhi-NCR Ghaziabad India Department of Computer Science and Business Systems Rajalakshmi Engineering College Thandalam India

Sentiment classification using emojis on social media has become increasingly crucial in recent years. Social media commonly uses emojis to convey feelings, emotions, and moods. Hence, in this article, a Jellyfish Algorithm based on a Deep Convolution Neural Network (JA-DCNN) is developed for the sentiment classification of the emojis. Initially, the emoji images are collected from the open-source system. The proposed sentiment classification process uses a Jellyfish Algorithm (JA) optimized Deep Convolution Neural Network (DCNN). In DCNN, the hyperparameters are selected with the help of JA to improve accuracy and efficiency. The proposed method aims to improve the training features, maintain the resolution, reduce memory consumption, and improve the system’s accuracy. The proposed method decreases the computation cost and memory consumption, which empowers the system’s performance in emojis-based sentiment classification. The proposed method is implemented in MATLAB, and performance metrics of similarity and quality measurements evaluate performances. To validate the performance of the proposed methodology, it is compared with conventional methods such as MobilenetV2, ResNet18, and ResNet50, which have been proven to obtain better performance metrics. © Bharati Vidyapeeth's Institute of computer Applications and Management 2025.

关键词： Classification Deep convolution neural network Emojis Jellyfish algorithm Sentiment analysis

来源：评论

学校读者我要写书评

暂无评论

Analysis of Ensemble Machine Learning Models for Fraud Detection

Analysis of Ensemble Machine Learning Models for Fraud Detec...

引用

2024 International Conference on Intelligent Systems for Cybersecurity, ISCS 2024

作者： Chaurasia, Siddharth Kesharwani, Satyam Sharma, Shanu Sharma, Swati Chugh, Bharti Kiet Group of Institutions Dept. of Computer Science and Engineering Ghaziabad India

ISBN: (纸本)9798350375237

Recent developments in electronic payment and e-commerce have led to a rise in digital fraud cases, including credit card fraud. Within the financial services industry, identifying credit card fraud is still a difficult task that can have significant effects on consumers' and financial institutions' reputations. Therefore, it is crucial to implement systems capable of detecting credit card fraud. This research utilizes the European Cardholder Data Collection to assess the efficacy of ensemble machine learning techniques in identifying credit card fraud. The European cardholder data set is subjected to a rigorous assessment and evaluation of series of machine learning models, such as Gradient Boosting, Random Forest and CAT Boost, to detect fraudulent transactions. It has been noted that the dataset is unbalanced, which may indicate that the models' performance is not very ideal. The study suggests using data sampling strategies to achieve a balanced distribution of data across various algorithms that yield optimal outcomes. The main goal is to identify the model that performs best for machine learning classification problems by comparing the performance of different models. The results of this study will support further initiatives to improve credit card transaction security and lower financial losses brought on by fraud. © 2024 IEEE.

关键词： Losses

来源：评论

学校读者我要写书评

暂无评论

Machine Learning based Malware Detection in Android: A Practical Approach 3

Machine Learning based Malware Detection in Android: A Pract...

引用

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

作者： Tyagi, Toshi Gautam, Madhu Kiet Group of Institutions Dept. of Computer Science and Engineering Ghaziabad India

ISBN: (纸本)9798350384598

Today's society heavily depends on online communities, which host a wide variety of social networks and public and private services. These demands led to the creation of Android, the world's most popular mobile operating system. The wide range of features and capabilities offered by freely installable third-party apps is largely responsible for this success. While focusing on the positive aspects of technology use, it is vital to keep in mind the potential negative consequences like privacy breaches, data theft, and malware infections that could compromise personal information and device security, especially if negligence or overuse of technology is present. This article discusses a systematic review of ML-based Android malware detection techniques. It provides a critical analysis of a number of carefully selected articles, highlighting their methods and conclusions. To support this study, we also present our own analysis and findings, for which an evaluation of diverse ML models with 123,453 applications and 5,560 malware samples from the DREBIN dataset was done to classify apps as malicious and benign. Preprocessing was done on the dataset to address missing values and perform label encoding. To cope with the issue of class imbalance, we used SMOTE an efficient ML technique that handles imbalance issues efficiently. We have applied five different ML algorithms, The results show that among all the techniques, Decision Tree (DT), Support Vector Machines (SVM) with various kernels (linear, polynomial, RBF), Sequential Neural Network (NN), K-Nearest Neighbours (KNN), and Logistic Regression (LR), the sequential neural network model outperformed other methods with 99% accuracy. Additionally, the process identifies areas for future research and improvement in ML-based Android malware detection techniques. © 2024 IEEE.

关键词： Android malware

来源：评论

学校读者我要写书评

暂无评论

Non-Invasive Skin Disease Diagnostic System Using Deep Learning 4

Non-Invasive Skin Disease Diagnostic System Using Deep Learn...

引用

4th International Conference on Technological Advancements in Computational sciences, ICTACS 2024

作者： Tomar, Vivek Bisht, Kamlesh Singh Panwar, Ayush Dimri, Reetesh Kumar Negi, Mayank Singh Sharma, Swati Computer Science & Engineering Department Uttarakhand Dehradun India Kiet Group of Institutions Computer Science & Engineering Department U.P. Ghaziabad India

ISBN: (纸本)9798350387490

Human skin is an important part of the body that needs attention. Skin diseases are quite common in a country like India with varied climate and high pollution levels. The usual diagnosis of skin diseases is based on visual examination and certain tests performed by a dermatologist. But this is a time-consuming process and can also end up in the wrong diagnosis. Skin disease diagnosis has undergone a revolution because of the significant development in the field of machine learning, notably convolutional neural networks (CNN). CNNs have the potential to completely change the way skin illnesses are diagnosed, tracked, and treated. This study investigates the application of various machine learning techniques in skin problem identification. By using CNN analysis of skin disease images, identification of patterns and making accurate predictions can be done with an unprecedented accuracy and consistency. In addition to diagnosis, machine learning models can treat different skins individually according to their type, thus simplifying the doctor's role. This study aims to use CNN to improve dermatology practices by introducing efficiency and accuracy. This study includes a comprehensive literature review examining existing CNN models and performance metrics. Goals include collecting health data, using CNN's predefined models, comparing their performance, recommending the best models for diagnosing skin diseases, and improving the user interface. It presents a benchmark-based neural network model that may contribute to advances in dermatology in identifying skin diseases through machine learning using data from HAM10000 and DERMNET. The effectiveness of the model is determined by comparing it with other existing models by evaluating quality indicators such as accuracy, precision, and F1 score. © 2024 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Streamlining Livestock Surveillance: YOLOv8 Approach to Cattle Body Segmentation

Streamlining Livestock Surveillance: YOLOv8 Approach to Catt...

引用

2024 International Conference on Intelligent Systems for Cybersecurity, ISCS 2024

作者： Chugh, Bharti Kumar, Pushepndra Mishra, Upendra Chaudhary, Himanshi Kiet Group of Institutions Dept. of Computer Science and Engineering Ghaziabad India

ISBN: (纸本)9798350375237

This research paper introduces an innovative approach to optimize livestock surveillance by employing YOLOv8 for cattle body segmentation. The study attained a mAP(50), mean Average Precision of 0.598 and 0.464 and mAP(50-95) on a dataset comprising 704 images. Preceding model training, extensive preprocessing and image augmentation techniques were implemented, resulting in a dataset of 1476 training, 171 validation, and 71 testing images. The developed model demonstrates considerable promise in enabling real-time mon-itoring of cattle and their health status. Additionally, to enhance its suitability for edge computing environments, quantization-aware training (QAT) abd weight compression techniques were employed, ensuring efficient deployment on resource-constrained devices. The results highlight the effectiveness and practicality of employing YOLOv8-based segmentation for livestock surveillance, with notable implications for precision agriculture and animal welfare. However, further enhancements are warranted, particularly in achieving higher mAP scores across a broader range of Intersection over Union (IoU) thresholds, to fully address the segmentation challenges in this domain. © 2024 IEEE.

关键词： computer vision

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：