检索结果-内蒙古大学图书馆

2024 International Conference on Knowledge Engineering and Communication Systems, ICKECS 2024

作者： Sai Akhil, M.C. Sanjana Sharma, B.S. Kodipalli, Ashwini Rao, Trupthi Global Academy of Technology Dept. of Artificial Intelligence & Data Science Karnataka Bangalore India

ISBN: (纸本)9798350359688

Medical reconstruction of images is essential for deep learning model training, improving diagnosis efficiency, and augmenting small datasets for a range medical purpose. In this work, we investigate the synthesis of realistic and superior chest X-ray pictures using Deep Convolutional neural networks Generative Adversarial Networks (DCGAN). A vital diagnosing technique for breathing problems disorders is chest X-ray imaging, and the development and assessment of AI-based medical devices can benefit from a greater amount of numerous, artificially produced information. We offer a novel method that creates synthetic chest Xray pictures that closely replicate real radiography data by utilizing DCGAN, a generative model architecture. Our solution is based on a two-step process: a discriminator network and a generator. The discriminator discerns among real and fake pictures, while a generator trains to produce accurate images of the chest. Through an intensive training regimen, these two networks continuously enhance the standard of the pictures they produce. We ran comprehensive tests on a variety of chest X-ray image datasets to confirm the efficacy of our methodology. Our DCGAN model produced artificial pictures that bore striking visual similarities to actual X-ray images of the chest, including radiographic noise, disease patterns, and anatomical features. In order to statistically evaluate the diversity and realism of the created photos, we also used a variety of assessment measures, which we used to show that our model created extremely convincing synthetic data. The synthesised chest X-ray pictures might be useful assets for expanding on small datasets, which will help build strong and precise deep-learning algorithms for the detection and classifications of pulmonary diseases. These pictures may also be utilized to improve computer-aided diagnosis system training and validation, which will help medical personnel make better judgements and provide better treatment

关键词： Pulmonary diseases

来源：评论

学校读者我要写书评

暂无评论

Machine Learning in Action: Supervised Learning Models for Classifying Credit Card Fraud

Machine Learning in Action: Supervised Learning Models for C...

引用

2024 International Conference on Electrical, Electronics and Computing Technologies, ICEECT 2024

作者： Pendalwar, Rishab Verma, Aditya Patil, Ratna Vishwakarma Institute of Information Technology Dept. of Artificial Intelligence and Data Science Pune India

ISBN: (纸本)9798350378092

Financial institutions face a relentless battle against fraudulent credit card transactions. Machine learning (ML) offers a promising approach for tackling this challenge. This paper explores six supervised machine learning models (Support Vector Machine (SVM), Random Forest, Naive Bayes, K-Nearest Neighbours (KNN), XGBoost, and LIGHT GBM) for detecting credit card fraud using a benchmark dataset. We explore the impact of dimensionality reduction with Principal Component Analysis (PCA) and data balancing with Generative Adversarial Networks (GANs) on imbalanced class distributions. The assessment relies on metrics such as accuracy, precision, recall, and F1-score. The results indicate that Random Forest demonstrates the best overall performance, especially when paired with GAN-based data balancing techniques. This study highlights the importance of addressing class imbalance and exploring dimensionality reduction techniques for enhancing credit card fraud detection capabilities. © 2024 IEEE.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Deep Transfer Learning Based Multi-Classification of Dental Diseases X-Ray Images 4

Deep Transfer Learning Based Multi-Classification of Dental ...

引用

4th International Conference on Ubiquitous Computing and Intelligent Information Systems, ICUIS 2024

作者： Narmadha, R.P. Karthik, B. Mithiloshini, B. Saranya, R. KIT-KalaignarKarunanidhi Institute of Technology Dept. Artificial intelligence and data science Coimbatore India

ISBN: (数字)9798331529635

ISBN: (纸本)9798331529635

This research work proposes a deep transfer learning model for multi-classification purposes within dental diseases classed by their X-ray images. The paper elaborates typical challenges posed by dentistry, which faces many diagnosis issues;the decisive process of proper classification into types of diseases requires being fine and efficient. Most state-of-the-art methodologies developed through CNNs and DNNs have been found to yield fair results, but they both suffer from overfitting and are time-consuming processes. The new model uses the ResNet50 to get over those limitations. Dental X-ray pictorials are to be grouped into four distinct categories using the proposed architecture: cavities, fillings, impacted teeth, and implants. The system proposed should classify these four categories in X-ray images with an accuracy greater than 92%. Therefore, the proposed system will automatically identify areas requiring clinical treatment. Expected outcome is a strong and dependable tool for enhanced dental disease diagnosis by better accuracy, efficiency, and performance in the evvaluation of X-ray pictures. © 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

Enhancing Student Engagement and Performance in Online Classrooms through a Smart Management Application 3

Enhancing Student Engagement and Performance in Online Class...

引用

3rd International Conference on Electronics and Renewable Systems, ICEARS 2025

作者： Gowthambalagi, M. Vishnu, S.M. NaveenKumar, M. Raja, S. K.S. Rangasamy College of Technology Department of Artificial Intelligence and Data Science Tiruchengode India

ISBN: (纸本)9798331509675

The fast pace of technological growth has revolutionized the education industry, and there is a need for embracing intelligent solutions to improve classroom management and student engagement. The Smart Classroom Management System (SCMS) takes advantage of innovative technologies like facial recognition, screen sharing, and real-time analytics to automate administrative processes, track student attentiveness, and promote interactive learning spaces. With process automation for features such as tracking attendance and evaluating student interaction, SCMS eliminates the strain on teachers but presents meaningful analytics for enhancing instructional techniques. Further, the software integrates AI-enabled components for one-to-one testing and adaptive questioning capabilities to promote tailored learning outcomes. SCMS allows hybrid learning and remote schooling by leveraging cloud capabilities, assuring accessibility, flexibility, and safety. This review research investigates current literature regarding smart classroom technologies, identifies major implementation challenges, and discusses future developments in AI-based education. The results focus on the use of SCMS to develop a dynamic and student-focused learning environment, ultimately enhancing educational outcomes and preparing students for the changing digital world. © 2025 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Comparison of Fair and FIFO Schedulers on Cloud Cluster of Apache Spark

Comparison of Fair and FIFO Schedulers on Cloud Cluster of A...

引用

2024 IEEE Conference on Engineering Informatics, ICEI 2024

作者： Reddy, Kasu Sai Kartheek Prasanna, R. Dhivya Shaik, Zuhair Hasan Thippireddy, Rishi Koushik Reddy Madhav, P. S. S. Vamsi Chaturvedil, Animesh Indian Institute of Information Technology Dharwad Department of Data Science and Artificial Intelligence Dharwad India

ISBN: (纸本)9798331505776

Apache Spark has become a key platform for processing large-scale datasets, particularly in predictive analytics and machine learning tasks, leveraging its in-memory capabilities and scalability. Efficient job scheduling is critical to optimising the performance of Spark clusters, minimising resource consumption, and reducing execution times. In this study, we compare the performance of two widely used Spark schedulers, FIFO and Fair, on Spark MLlib tasks for two algorithms, Parallel Frequent Pattern Growth (PFP-Growth) and Random Forest, focusing on their execution times and resource utilisation across varying cloud cluster sizes. To perform this analysis, we set up Spark clusters with different configurations, ranging from 1 to 30 worker nodes, in a controlled environment. Our experiments reveal that the Fair scheduler consistently outperforms the FIFO scheduler, particularly as the cloud cluster scales from small to moderate sizes. The Fair scheduler's ability to dynamically allocate resources and manage concurrent jobs more efficiently leads to reduced execution times and better overall cloud cluster utilisation, making it a preferable choice for multi-job Spark workloads. These findings offer insights to optimise Spark deployments in real-world applications, where resource management and task scheduling play crucial roles in performance and cost efficiency. © 2024 IEEE.

关键词： Cloud Computing Cluster Fair scheduler FIFO scheduler Job scheduling Spark MLlib

来源：评论

学校读者我要写书评

暂无评论

Predictive Modeling for Concrete Compressive Strength: ML-Regressors, Custom Neural Networks, and Explainable AI Approaches 4

Predictive Modeling for Concrete Compressive Strength: ML-Re...

引用

4th IEEE Asian Conference on Innovation in technology, ASIANCON 2024

作者： Bharath, M. Gowtham, S. Kodipalli, Ashwini Rao, Trupthi Global Academy of Technology Dept. of Artificial Intelligence & Data Science Karnataka Bangalore India

ISBN: (纸本)9798350354218

Concrete compressive strength is crucial for optimizing concrete mix designs in construction. This study employs traditional machine learning regressors, custom ANNs, and Explainable AI (XAI) techniques using a dataset from Kaggle with 1030 instances and 9 attributes. Among the models tested, Random Forest Regression exhibited the best performance with the lowest MSE and highest R-squared (R2) values. Custom ANNs, particularly a five-layer architecture with SeLU and ReLU activations optimized using the Adam optimizer, achieved superior results with an MSE of 22.2 and an R2 of 0.92. To ensure model transparency, we used the LIME technique, revealing the contributions of individual input variables. Our findings highlight the effectiveness of advanced ML models and XAI techniques in predicting concrete compressive strength, offering a robust framework for optimizing concrete mix designs and enhancing construction practices. Future research can expand on these results by exploring larger datasets, additional features, and more advanced XAI methods. © 2024 IEEE.

关键词： Decision trees

来源：评论

学校读者我要写书评

暂无评论

Traffic Scene Text Detection for 3D Environment 12

Traffic Scene Text Detection for 3D Environment

引用

12th International Conference on Current Trends in Advanced Computing, ICCTAC 2024

作者： Sarumathi, S. Geetha Chawla, M. Manjusri, S. Renukadevi, K. K S Rangasamy College of Technology Department of Artificial Intelligence and Data Science Tamilnadu India

ISBN: (数字)9798350395808

ISBN: (纸本)9798350395808

With an emphasis on text localization, extraction, and recognition, this project outlines the development of an optical character recognition (OCR) system that makes use of the OpenCV (Open-Source Computer Vision) library. Text may be extracted from images, recognize it with OCR, and identify text portions utilizing OpenCV's image processing and analysis capabilities. With applications ranging from document digitization to image text extraction and data entry automation, the project's adaptability and accessibility make it a priceless tool for automating text-related tasks. This makes it a good choice for companies, researchers, and developers interested in OCR technology. By using OpenCV, the system's capabilities are enhanced, making it a more reliable and adaptable method for text extraction from images. © 2024 IEEE.

关键词： Optical character recognition

来源：评论

学校读者我要写书评

暂无评论

Efficient segmentation of exudates in color fundus images using wavelets and generative adversarial network

引用

Multimedia Tools and Applications 2025年 1-31页

作者： Kar, Mithun Kumar Nath, Malaya Kumar Neog, Debanga Raj Amrita School of Artificial Intelligence Coimbatore Amrita Vishwa Vidyapeetham Coimbatore641112 India Department of Electronics and Communication Engineering National Institute of Technology Puducherry Karaikal609609 India Mehta Family School of Data Science and Artificial Intelligence Indian Institute of Technology Guwahati Guwahati781039 India

Due to the progression of diabetic retinopathy (DR), retinal features change and lead to the appearance of microaneurysms, hemorrhages, and exudates on its surface. The presence of exudates appears over the retina, giving a clear early sign of diabetic retinopathy, which is an irreversible eye disease. Automatic segmentation and detection of exudates enhance the computer-aided diagnosis of the disease. Various deep learning techniques are employed for the automatic detection and segmentation of exudates, but their performance is unsatisfactory due to the small size of exudates. Here, a novel method is proposed, using generative adversarial network on wavelet subbands. Wavelet decomposition un-correlates exudates from other retinal features in subbands due to its frequency-selective nature. Generative adversarial network (GAN) is employed with modified loss function with multi-scale residual convolutional neural network (MSR-Net) to accurately segment exudates. The proposed method is experimented on using IDRiD and e-optha databases, and different quantitative and qualitative performance measures are computed and compared with state-of-the-art methods. The proposed method achieves an accuracy, sensitivity, and F1Score of 0.9886, 0.8306, and 0.8246, respectively, for the IDRiD database. For the e-ophtha database, the above-mentioned performance measures are reported to be 0.9861, 0.8512, and 0.8444, respectively. This method performed well in presence of other features. For generalization purposes, the proposed method is applied for skin cancer cell segmentation using University of Waterloo’s skin cancer database. It achieved highest accuracy of 0.9943 and sensitivity of 0.9885 for segmenting the skin cancer cells. It has the potential for analyzing the diagnostic information about various ophthalmic diseases such as: glaucoma and diabetic retinopathy (DR). © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2025.

关键词： Wavelet decomposition

来源：评论

学校读者我要写书评

暂无评论

BSDIB: Balanced Self-Distillation Information Bottleneck for Long-Tailed Visual Recognition 2

BSDIB: Balanced Self-Distillation Information Bottleneck for...

引用

2nd International Conference on Frontiers of artificial intelligence and Machine Learning, FAIML 2023

作者： Cai, Xin Tan, Shan School of Artificial Intelligence and Automation Huazhong University of Science and Technology China

ISBN: (纸本)9798400707544

Real-world data are usually long-tailed, which results in poor performance of visual recognition systems. Current prevailing methods decouple the training procedure into representation learning and classification. However, the two stages in such methods are independent, resulting in sub-optimal solutions. In this paper, we propose a one-stage method, namely balanced self-distillation information bottleneck (BSDIB), to take care of both representation learning and classification simultaneously. Specifically, a re-weighting strategy is used to obtain an unbiased classifier, and the class-dependent self-distillation information bottleneck is utilized to promote data representations. Extensive experiments on three commonly-used long-tailed datasets indicate that BSDIB achieves SOTA performance and the visualization results show that our approach focuses more on the target object. © 2023 Copyright held by the owner/author(s).

关键词： Distillation

来源：评论

学校读者我要写书评

暂无评论

Enhancing Alpha Thalassemia Screening: A Comparative Study of Multiple Machine Learning Classifiers and Interpretation using Explainable AI 4

Enhancing Alpha Thalassemia Screening: A Comparative Study o...

引用

4th International Conference on Intelligent Technologies, CONIT 2024

作者： Bharath, M. Gowtham, S. Kodipalli, Ashwini Rao, Trupthi Global Academy of Technology Dept. of Artificial Intelligence & Data Science Karnataka Bangalore India

ISBN: (纸本)9798350349900

This study introduces an innovative paradigm for Alpha thalassemia detection by integrating machine learning, hyperparameter tuning, and explainable artificial intelligence (XAI) techniques. Through a comparative analysis, boosting algorithms, especially XG Boost, demonstrate remarkable efficacy in accurately identifying Alpha thalassemia conditions. The iterative hyperparameter tuning process enhances the final XG Boost model's accuracy to an impressive 90.11%. The incorporation of XAI methods, such as SHAP values and LIME, provides a nuanced understanding of feature importance and individual predictions, elevating model interpretability. This research's significance lies in its potential to transform Alpha thalassemia screening, offering precision, resilience, and transparency. The amalgamation of machine learning classifiers, hyperparameter tuning, and XAI techniques contributes to advanced screening tools, fostering trust among clinicians and researchers. © 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

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：