检索结果-内蒙古大学图书馆

28th International computer science and Engineering Conference, ICSEC 2024

作者： Janprasit, Siwach Punkong, Narong Ratanavilisagul, Chiabwoot Kosolsombat, Somkiat Faculty of Applied Science Department of Computer and Information Science Bangkok Thailand Digital Technology for Business Faculty of Management Science Kanchanaburi Thailand Faculty of Applied Science Department of Computer and Information Sciences Bangkok Thailand Thammasat University Data Science and Innovation College of Interdisciplinary Studies Thailand

ISBN: (纸本)9798350366860

handwritten digit recognition is a crucial task in various fields such as postal mail sorting, bank check processing, and digitizing handwritten documents. This research aims to compare the effectiveness of using Convolutional Neural Network (CNN) for feature extraction combined with Support Vector Machine (SVM) for classification, against the traditional method of using Discrete Wavelet Transform (DWT) and Discrete Cosine Transform (DCT) for feature extraction. Additionally, Particle Swarm Optimization (PSO) is applied to optimize SVM parameters to further enhance model performance. The experiments are conducted on two datasets: The MNIST dataset and an individual handwriting dataset from 50 participants. The results indicate that the proposed method of using CNN for feature extraction and PSO-optimized SVM achieves superior accuracy compared to traditional DWT and DCT methods. Specifically, the proposed method achieved an accuracy of 99.05% on the MNIST dataset and 67.79% on the individual handwriting dataset, demonstrating its robustness and effectiveness in handwritten digit recognition. © 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

Optimizing classifier performance using PCA-FLANN: a fast and reliable approach

引用

Iran Journal of computer science 2025年第2期8卷 457-470页

作者： Jena, Monalisa Behera, Ranjan Kumar Dehuri, Satchidananda Cho, Sung-Bae Department of Computer Science Yonsei University Seoul 03722 South Korea Department of Computer Science & Engineering Indian Institute of Information Technology Jharkhand Ranchi India Department of Computer Science Fakir Mohan University Odisha Balasore India

This work introduces PCA-FLANN, an innovative hybrid model combining principal component analysis (PCA) with functional link artificial neural network (FLANN) to achieve efficient non-linear dimensionality reduction and robust classification. The approach is motivated by the need to optimize feature selection, recognizing that not all features equally contribute to achieving optimal accuracy. PCA is employed to extract the most relevant features from datasets, retaining essential information while significantly reducing dimensionality, and FLANN processes these reduced features to handle non-linear correlations effectively. To assess its performance, PCA-FLANN is evaluated on six widely-used real-world datasets from the KEEL repository and compared against hybrid models such as PCA-RBFN, PCA-SVM, PCA-RF, PCA-MLP, and standalone FLANN. The results reveal that PCA-FLANN outperforms these models in accuracy, with improvements of 8.5%, 2.94%, 9.09%, 11.1%, and 9.69%, respectively. Additionally, the model demonstrates enhanced computational efficiency and reduced complexity, establishing PCA-FLANN as a powerful and precise solution for classification tasks involving dimensionality reduction. © The Author(s), under exclusive licence to Springer Nature Switzerland AG 2025.

关键词： Classification Feature extraction FLANN Neural network PCA

来源：评论

学校读者我要写书评

暂无评论

Building an effective KPI monitoring system for an OTT Platform 4

Building an effective KPI monitoring system for an OTT Platf...

引用

4th International Conference on Emerging Systems and Intelligent Computing, ESIC 2024

作者： Jena, Biswaranjan Mishra, Debahuti Mishra, Smitaprava University Department of Computer Science and Engineering Bhubaneswar India University Department of Computer Science and Information Technology Bhubaneswar India

ISBN: (纸本)9798350349856

Success of any Over the Top (OTT) platform depends on how the platform is providing best user experience along with content to its customers. Being in media and entertainment space customers need to access content from various devices like mobile, tab, desktop, tv including set top boxes. To measure the effectiveness of the systems its inevitable to define standard Key Performance Indicators (KPIs) and continuously monitor the application along with standard monitoring capabilities at infrastructure level to make the application available with required throughput. This paper describes in detail various application KPIs respective to different channels and systems along with tools used to measure the same. © 2024 IEEE.

关键词： Benchmarking

来源：评论

学校读者我要写书评

暂无评论

Speaker diarization based on X vector extracted from time-delay neural networks (TDNN) using agglomerative hierarchical clustering in noisy environment

引用

International Journal of Speech technology 2025年第1期28卷 13-26页

作者： Khadar, K. V. Aljinu Sunil Kumar, R.K. Sameer, V.V. Department of Computer Science University of Calicut Tenhippalam Kerala Malappuram673635 India Department of Information Technology Kannur University Mangattuparamba Kerala Kannur670567 India Department of Computer Science Farook College Kerala Kozhikode673 632 India

This paper introduces a speaker diarization system using speaker embedding parameters, specifically the x-vector. By incorporating auto-correlated MFCC features for x-vector extraction using a pre-trained time delay neural network, the system exhibits enhanced adaptability to noise variations. Speaker clustering is accomplished through agglomerative clustering with PLDA scoring as the distance metric, making the system particularly valuable for potential speaker identification, especially in forensic applications. The system’s noise adaptability is thoroughly evaluated by integrating various types of noise such as Red, Pink, and white noise across a wide range of signal-to-noise ratios (20 dB to − 20 dB). Additionally, the system’s performance is comprehensively assessed by varying speech duration and adjusting the number of speakers, highlighting its robustness and effectiveness in real-world scenarios. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2024.

关键词： Vectors

来源：评论

学校读者我要写书评

暂无评论

Towards smart technologies with integration of the internet of things, cloud computing, and fog computing

引用

International Journal of Networking and Virtual Organisations 2023年第1期29卷 73-124页

作者： Ahlawat, Chanchal Krishnamurthi, Rajalakshmi Department of Computer Science and Engineering Jaypee Institute of Information Technology Noida201309 India

In the recent era of technology, the internet of things (IoT) plays a tremendous role in enhancing the quality of human life through smart devices and sensing the real-world environment. IoT aims to interconnect anything from anywhere via the internet. The existing literature lacks in providing deeper research insight toward a broader perspective of integrating IoT, cloud computing (CC), and fog computing (FC). Hence, this paper aims to provide a comprehensive review of integrating these main emerging technologies and four major objectives are addressed. The first objective identifies the primary characteristics of IoT to solve real-life problems. The second objective addresses the issues that necessitate the integration of IoT with other emerging technologies. The third objective involves how the integration of IoT with CC elevates the scope of providing a solution to real-life problems. The final objective addresses the limitations of the integration of IoT with CC and how FC overcomes these limitations. © 2023 Inderscience Enterprises Ltd.

关键词： Internet of things

来源：评论

学校读者我要写书评

暂无评论

Data Augmentation and Random Multi-Model Deep Learning for Data Classification

引用

computers, Materials & Continua 2023年第3期74卷 5191-5207页

作者： Fatma Harby Adel Thaljaoui Durre Nayab Suliman Aladhadh Salim EL Khediri Rehan Ullah Khan Computer Science Department Future Academy-Higher Future Institute for Specialized Technological StudiesEgypt Department of Computer Systems Engineering Faculty of Electrical and Computer EngineeringUniversity of Engineering and TechnologyPeshawar25120Pakistan Department of Information Technology College of ComputerQassim UniversityBuraydahSaudi Arabia Department of Computer Sciences Faculty of Sciences of GafsaUniversity of GafsaGafsaTunisia

In the machine learning(ML)paradigm,data augmentation serves as a regularization approach for creating ML *** increase in the diversification of training samples increases the generalization capabilities,which enhances the prediction performance of classifiers when tested on unseen *** learning(DL)models have a lot of parameters,and they frequently ***,to avoid overfitting,data plays a major role to augment the latest improvements in ***,reliable data collection is a major limiting ***,this problem is undertaken by combining augmentation of data,transfer learning,dropout,and methods of normalization in *** this paper,we introduce the application of data augmentation in the field of image classification using Random Multi-model Deep Learning(RMDL)which uses the association approaches of multi-DL to yield random models for *** present a methodology for using Generative Adversarial Networks(GANs)to generate images for data *** experiments,we discover that samples generated by GANs when fed into RMDL improve both accuracy and model *** across both MNIST and CIAFAR-10 datasets show that,error rate with proposed approach has been decreased with different random models.

关键词： Data augmentation generative adversarial networks classification machine learning random multi-model deep learning

来源：评论

学校读者我要写书评

暂无评论

A transformer-based Urdu image caption generation

引用

Journal of Ambient Intelligence and Humanized Computing 2024年第9期15卷 3441-3457页

作者： Hadi, Muhammad Safder, Iqra Waheed, Hajra Zaman, Farooq Aljohani, Naif Radi Nawaz, Raheel Hassan, Saeed Ul Sarwar, Raheem Department of Computer Science National University of Computer & ampEmerging Sciences Lahore Pakistan Department of Computer Science Information Technology University Lahore Pakistan Faculty of Computing and Information Technology King Abdulaziz University Jeddah Saudi Arabia Staffordshire University Stoke-on-Trent United Kingdom Department of Computing and Mathematics Manchester Metropolitan University Manchester United Kingdom Faculty of Business and Law Manchester Metropolitan University Manchester United Kingdom

Image caption generation has emerged as a remarkable development that bridges the gap between Natural Language Processing (NLP) and computer Vision (CV). It lies at the intersection of these fields and presents unique challenges, particularly when dealing with low-resource languages such as Urdu. Limited research on basic Urdu language understanding necessitates further exploration in this domain. In this study, we propose three Seq2Seq-based architectures specifically tailored for Urdu image caption generation. Our approach involves leveraging transformer models to generate captions in Urdu, a significantly more challenging task than English. To facilitate the training and evaluation of our models, we created an Urdu-translated subset of the flickr8k dataset, which contains images featuring dogs in action accompanied by corresponding Urdu captions. Our designed models encompassed a deep learning-based approach, utilizing three different architectures: Convolutional Neural Network (CNN) + Long Short-term Memory (LSTM) with Soft attention employing word2Vec embeddings, CNN+Transformer, and Vit+Roberta models. Experimental results demonstrate that our proposed model outperforms existing state-of-the-art approaches, achieving 86 BLEU-1 and 90 BERT-F1 scores. The generated Urdu image captions exhibit syntactic, contextual, and semantic correctness. Our study highlights the inherent challenges associated with retraining models on low-resource languages. Our findings highlight the potential of pre-trained models for facilitating the development of NLP and CV applications in low-resource language settings. © The Author(s) 2024.

关键词： Long short-term memory

来源：评论

学校读者我要写书评

暂无评论

Wavelet Based Classification Using Meta-Heuristic Algorithm with Deep Transfer Learning Technique

引用

SN computer science 2025年第3期6卷 1-25页

作者： Pandey, Anupam Pandey, Vikas Kumar Department of Computer Science and Information Technology Guru Ghasidas Vishwavidyalaya Chahttisgarh Bilaspur India

Early identification of brain tumors is crucial for cancer diagnosis since it can greatly increase survival chances. Brain tumors, which are defined as abnormal cell growth inside the brain, are among the worst types of cancer. Accurate diagnosis and effective treatment of brain cancers heavily rely on medical imaging. However, early detection is inherently challenging due to the elusive nature of tumors. To analyze malignancies, magnetic resonance imaging (MRI) pictures of patients are crucial. Due to several misclassifications, manual procedures have been unable to keep up with the volume of created data. With a deep transfer learning method, five primary working modules are used in the suggested research work to deal with huge amounts of data and reduce miss-classification first, the tumor pictures are stripped of their skulls, Afterward, the skull-stripped images are filtered using the guided bilateral filter (GBF). Subsequently, the areas affected by brain tumors are segmented using the OTSU optimum thresholding approach and transformed utilizing the wavelet-based reconstruction technique. The primary features of the MRI image are extracted using the improved Gabor wavelet Transform (IGWT) technique. The Grey Wolves Optimization (GWO) method selects the best features, which are then supplied to the deep-transfer-learning classification model. In this paper, we employed five deep transfer learning architectures: VGG19, InceptionV3, InceptionResNetV2, ResNet152, and DenseNet121. A dataset that was assembled from three benchmark databases used in our research. DenseNet121 deep transfer learning model works better than other techniques, as demonstrated by the testing results, which show an accuracy of 99.43%. © The Author(s), under exclusive licence to Springer Nature Singapore Pte Ltd. 2025.

关键词： Deep Transfer Learning Feature extraction MRI imaging OTSU threshold Wavelet Transformation

来源：评论

学校读者我要写书评

暂无评论

Malware Detection Based on Stack Ensemble Using Machine Learning 2

Malware Detection Based on Stack Ensemble Using Machine Lear...

引用

2nd International Conference on Computing and Data Analytics, ICCDA 2024

作者： Eke, Christopher Ifeanyi Olamide, Fawaz Ibrahim Phiri, Musa Mwenge, Mulenga Faith Fatokun, B. Shuib, Liyana Federal University of Lafia Faculty of Computing Department of Computer Science Nasarawa State Nigeria Mulungushi University Department of Computer Science and Information Technology Kabwe Zambia Malaysian Institute of Information Technology University of Kuala Lumpur Kuala Lumpur Malaysia University of Malaya Faculty of Computer Science & It Department of Information Systems Kuala Lumpur Malaysia

ISBN: (纸本)9798331542573

information security remains one of the major challenges faced by organizations and individuals in the current technological era. With the growing popularity of smart devices, the frequency of cyber-attacks targeted at these devices are increasing at a rapid pace. Among these types of attacks on information systems is malicious software also known as malware. A cyber-attack based on malware has the potential to cause loss or manipulation of confidential information. Currently, malware creators who use polymorphic techniques to evade detection can bypass traditional signature-based methods of detecting malware. Thus, various researchers have turned to Machine Learning (ML) techniques to enhance the detection of malware. However, despite the promising results, ML-based classification of malware remains a challenging task that the research community is yet to solve. Therefore, in this study, we propose an ML-based malware detection system that incorporates the AdaBoost, Random Forest, Decision Tree, and K-Nearest Neighbors Algorithm (KNN) algorithms into an ensemble stack with gradient boost as the ensemble boost for the stack meta estimator. We also use Extra Trees Classifier for feature selection. Multiple performance measures are utilized to assess the system's performance while using the Kaggle Malware Dataset. Our findings demonstrate that the proposed method successfully detects malware samples with remarkable accuracy, precision, recall, F1, and area under the curve (AUC) scores of 99.4%, 99.3%, 99.4%, 99.4%, and 99.6%, respectively. © 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

Deep Learning-Driven Edge-Enabled Serverless Architectures for Animal Emotion Detection

Informatica (Slovenia)

引用

Informatica (Slovenia) 2025年第7期49卷 33-48页

作者： Benedict, Shajulin Subair, Rubiya Department of Computer Science and Engineering Indian Institute of Information Technology Kottayam Kerala India Technical University Munich Garching Germany

Animal emotion detection, including elephant emotions, is highly possible, but what the traditional emotion detection approaches highlight is their blatant ignorance of adopting edge-enabled intelligence and serverless-based solutions, both of which are affordable. Treating the emotions of animals increases their productivity, especially among trained elephants when subjected to carrying logs or undertaking gargantuan tasks. However, existing infrastructures are inefficient in handling long-running animal emotion detection-related tasks. This article proposes a deep learning-driven edge-enabled serverless architecture after evaluating several existing animal emotion detection techniques. Additionally, we perform an exploratory study on the cost impact of incorporating serverless-enabled approaches to animal emotion detection architectures. We observed that the proposed edge-enabled serverless architectures saved over 13,000 dollars annually compared to traditional animal emotion detection approaches. In addition, the article provided a few research directions to develop novel edge-enabled serverless architectures that boost socio-economic situations while avoiding human-animal conflicts. © 2025 Slovene Society Informatika. All rights reserved.

关键词： Invertebrates

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：