检索结果-内蒙古大学图书馆

Feature Selection with Deep Belief Network for Epileptic Seizure Detection on EEG Signals

Computers, Materials & Continua 2023年第5期75卷 4101-4118页

作者： Srikanth Cherukuvada R.Kayalvizhi Department of Networking and Communications School of ComputingSRM Institute of Science and TechnologyKattankulathurChennai603203India

The term Epilepsy refers to a most commonly occurring brain disorder after a *** identification of incoming seizures significantly impacts the lives of people with *** detection of epileptic seizures(ES)has dramatically improved the life quality of the *** Electroencephalogram(EEG)related seizure detection mechanisms encountered several difficulties in *** EEGs are the non-stationary signal,and seizure patternswould changewith patients and recording ***,EEG data were disposed to wide noise varieties that adversely moved the recognition accuracy of *** intelligence(AI)methods in the domain of ES analysis use traditional deep learning(DL),and machine learning(ML)*** article introduces an Oppositional Aquila Optimizer-based Feature Selection with Deep Belief Network for Epileptic Seizure Detection(OAOFS-DBNECD)technique using EEG *** primary aim of the presented OAOFS-DBNECD system is to categorize and classify the presence of *** suggested OAOFS-DBNECD technique transforms the EEG signals *** format at the initial ***,the OAOFS technique selects an optimal subset of features using the preprocessed *** seizure classification,the presented OAOFS-DBNECD technique applies Artificial Ecosystem Optimizer(AEO)with a deep belief network(DBN)*** extensive range of simulations was performed on the benchmark dataset to ensure the enhanced performance of the presented OAOFS-DBNECD *** comparison study shows the significant outcomes of the OAOFS-DBNECD approach over other *** addition,the result of the suggested approach has been evaluated using the CHB-MIT database,and the findings demonstrate accuracy of 97.81%.These findings confirmed the best seizure categorization accuracy on the EEG data considered.

关键词： Seizure detection EEG signals machine learning deep learning feature selection

来源：评论

学校读者我要写书评

暂无评论

YOLOv8-Based Model for Precise Corrosion Segmentation in Industrial Imagery

YOLOv8-Based Model for Precise Corrosion Segmentation in Ind...

引用

2024 International Conference on Computational Intelligence for Green and Sustainable technologies, ICCIGST 2024

作者： Chandu, Seethala Devi Varma, Sai Ranjani, M. SRM Institute of Science and Technology Kattankulathur Department of Computing Technologies Chennai India SRM Institute of Science and Technology Kattankulathur Faculty of Engineering and Technology Department of Computing and Technologies Chennai India

ISBN: (纸本)9798350381887

A significant risk to the longevity and intelligence of fundamental infrastructure, such bridges and pipelines, is corrosion. Easy finding and erosion monitoring are important for predicting additional setbacks and ensuring transparency in security. This investigation offers a creative method of erosion discovery by utilizing the YOLOv8 question discovery computation. It also works in tandem with the WANDB stage to facilitate efficient show planning and organization. Said to offer a new method of managing videos by combining the powerful visualization powers of the W&B stage with YOLOv8, an enhanced version of the YOLO protest demonstration. YOLOv8 refers to a state-of-the-art development within the YOLO family, intended to surpass expectations in real-time protest finding tasks with increased productivity and accuracy. Additionally, considered looking into the use of cloud-based virtual machines as an affordable alternative to traditional setups like purplish blue for show organization and inspection, with a focus on the W&B Cloud advertising. One of YOLOv8's unique features is its single-pass localization method, which handles the entire video outline in a single forward pass over the neural network. This increases the pace of discovery and eliminates the need for intricate post-processing. Improved question localization and classification accuracy are encouraged by secured boundary boxes and network cell designs. The real-time position optimization of YOLOv8's engineering makes it particularly suitable for video feeds. Furthermore, by employing the Weights & Predispositions stage, analysts and experts can visualize and monitor the performance of the YOLOv8 demonstration during video preparation. W&B provides easily navigable dashboards and visualizations that enable real-time tracking of important metrics, demonstrate behavior, and pinpoint the site of protests. This integration provides a deeper comprehension of the model's features and potential areas of improv

关键词： Demonstrations

来源：评论

学校读者我要写书评

暂无评论

Efficient Rent Price Prediction Model for the Development of a House Marketplace Website by Analyzing Various Regression-Based Machine Learning Algorithms 7th

Efficient Rent Price Prediction Model for the Development of...

引用

7th International Conference on Inventive Communication and Computaional technologies, ICICCT 2023

作者： Saraswat, Ojas Arunachalam, N. Department of Computing Technologies School of Computing SRM Institute of Science and Technology Chennai India

ISBN: (纸本)9789819951659

Nowadays, as people move to new cities, renting a house according to the needs at affordable price has become a necessity. To solve this problem efficiently, this paper develops and shows the implementation of a complete house marketplace website that uses an effective rent price prediction model to help both customers and owner buy/rent property at accurate prices. For the dataset, which holds details of more than 22,000 properties in Pune, India, multiple regression-based machine learning (ML) algorithms have been used on various parameters to determine the rent prices and figure out which algorithm performs best. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

Brain Tumor Detection Using Deep Learning: ResNet+LSTM 5

Brain Tumor Detection Using Deep Learning: ResNet+LSTM

引用

5th International Conference on IoT Based Control Networks and Intelligent Systems, ICICNIS 2024

作者： Balaji, Sriraam Krithika, T. Joshi, Soshya SRM Institute of Science and Technology Department of Computing Technologies Chennai India

ISBN: (纸本)9798331518097

Brain tumors are among the most severe medical conditions, necessitating early and accurate detection to improve patient outcomes. This research paper explores a hybrid deep learning model that combines. Combining Long Short-Term Memory (LSTM) models with Residual Networks (ResNet-50) improves the ability to identify brain cancers from MRI data. The Brain MRI Images for Tumor Detection dataset, which is accessible on FigShare, contains annotated MRI scans of a variety of tumor forms, including pituitary tumors, meningiomas, and gliomas. This study makes use of this dataset. For the purpose of comprehending the evolution of brain tumors, the model uses ResNet-50 to extract spatial characteristics from individual MRI slices, while LSTM records temporal relationships across several slices. The dataset comprises images that were preprocessed by resizing to 128×128 pixels and normalizing pixel values for con-sistent input. The proposed ResNet-LSTM model classifies MRI slices into three categories: gliomas, meningiomas and pituitary. Performance metrics such as accuracy, precision, and confusion matrix were used to evaluate the model, demonstrating significant improvements over traditional methods. The findings suggest that the integration of ResNet and LSTM can effectively capture both spatial and sequential information, enhancing diagnostic accuracy and reliability. The potential of hybrid deep learning models to improve medical imaging diagnoses is highlighted in this study. © 2024 IEEE.

关键词： Magnetic resonance imaging

来源：评论

学校读者我要写书评

暂无评论

Detection and Classification of Brain Tumor Using Deep Learning Model 1

引用

4th International Conference on Computer Vision, High-Performance computing, Smart Devices, and Networks, CHSN 2023

作者： Routhu, Manoj Sitaram Govindaraju, Rohith Vigna Saranya, P. Department of Computing Technologies School of Computing College of Engineering and Technology SRM Institute of Science and Technology Kattankulathur Chennai603203 India

ISBN: (数字)9789819777945

ISBN: (纸本)9789819777938

Accurate and efficient brain tumor classification is paramount for timely clinical diagnosis and effective treatment planning. In this groundbreaking research, we introduce an innovative Convolutional Neural Network (CNN) architecture intricately integrated with customized preprocessing techniques, resulting in an exceptional classification accuracy of 98.5% on the challenging br35H dataset. By harnessing the power of MRI scans and leveraging diverse datasets, our model expedites brain tumor assessments and sets the stage for advanced classification methodologies. With the global incidence of brain tumors rising, the need for technology-driven diagnosis becomes increasingly evident, and CNNs emerge as pivotal tools in enhancing diagnostic precision. This study not only underscores the profound significance of CNN models but also transcends geographical boundaries, reducing the frequency of misdiagnoses and ultimately empowering global healthcare. This paper comprehensively explores our methodology, delving into the intricate details of data collection processes, model development strategies, and experimental findings. Moreover, it sheds light on the implications of deploying CNN models in medical imaging. By contributing to the ongoing discourse on transformative healthcare technologies, this research aims to propel the adoption of CNN-based approaches, ushering in a new era of precise and efficient brain tumor classification for the benefit of healthcare professionals, patients, and society. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Agriconnect: A Data-Driven Platform for Optimizing Crop Production with Fine-Tuned Large Language Models 10

Agriconnect: A Data-Driven Platform for Optimizing Crop Prod...

引用

10th International Conference on Electrical Energy Systems, ICEES 2024

作者： Jeya, R. Singh, Dimpy Jha, Anuj Mallick, Soumik SRM Institute of Science and Technology Department of Computing Technologies Chennai India

ISBN: (纸本)9798350353778

Agriconnect, a web-based platform, empowers farmers with data-driven crop production insights through a two-model machine learning approach. The Crop Insights Model leverages a fine-tuned large language model (LLM) called Gemini. Gemini is first enriched with dealer information from a user-created, state-categorized CSV file. To ensure location accuracy, SerpApi retrieves data from Google Maps based on user searches, which is then fed into a vector database alongside the dealer information. This comprehensive dataset fine-tunes Gemini for location-specific insights. The Crop Recommender Model then utilizes the enhanced Gemini to analyze user-provided climate and soil data (type and moisture). Based on this analysis, the fine-tuned LLM recommends crops best suited for specific conditions, aiming to optimize crop selection and mitigate climate and market fluctuations risks. Finally, an interactive visualization dashboard presents this information, empowering farmers to make informed decisions and implement effective practices. © 2024 IEEE.

关键词： Data accuracy

来源：评论

学校读者我要写书评

暂无评论

Light Weight CNN Based on Knowledge Distillation for Diabetic Retinopathy Detection 4

Light Weight CNN Based on Knowledge Distillation for Diabeti...

引用

4th International Conference on Expert Clouds and Applications, ICOECA 2024

作者： Baranidharan, B. Janenie, J. Chhipa, Hardik SRM Institute of Science and Technology Department of Computing Technologies Kattankulathur India

ISBN: (纸本)9798350385793

Diabetic Retinopathy (DR) is a severe consequence of diabetes, which poses a significant threat to vision, potentially leading to irreversible blindness if not diagnosed in its early stages. It generally impacts 80% of people, who have had diabetes for more than ten years. This condition arises from diabetes-induced damage to retinal blood vessels, emphasizing the critical need for precise and timely intervention and diagnosis. Traditional screening methods depend on the manual examination of colored fundus images by trained specialists;there are chances of errors that might miss the early signs of diabetic retinopathy. Various Convolutional Neural Networks (CNNs) such as DenseNet121, DenseNet169, DenseNet201, MobileNet, ResNet152, Xception, and Inception are explored to address this challenge. It is engaged with these models through rigorous experimentation to examine their capabilities to enhance the precision of Diabetic Retinopathy (DR) diagnosis in all its stages. Though these models give efficient accuracies, the size of these models makes it complex to be embedded in miniature medical devices. This study has introduced the application of knowledge distillation through implementing the teacher-student model. This research study aims to achieve a lighter, more compact, and simpler student model, which is derived from the more complex teacher model, which, in this case, is the aforementioned model. After thorough analysis, this study found that the student model mentored by the DenseNet169 outperforms other models in accurately detecting Diabetic Retinopathy (DR) with the highest validation accuracy of 68. 77%. Using a diverse dataset from Kaggle, our focus goes beyond improving accuracy. This study aims to contribute valuable insights to fight against diabetes complication. © 2024 IEEE.

关键词： Diagnosis

来源：评论

学校读者我要写书评

暂无评论

Leveraging AI for Dyslexia Detection: Exploring Multimodal Approaches with Machine Learning and Deep Learning 4

Leveraging AI for Dyslexia Detection: Exploring Multimodal A...

引用

4th International Conference on Sentiment Analysis and Deep Learning, ICSADL 2025

作者： Priyasri, G.R. Devi, M. Uma Srm Institute of Science and Technology Department of Computing Technologies Kattankulathur Chengalpattu India

ISBN: (纸本)9798331523923

Developmental dyslexia is a learning disorder that typically begins in early childhood, making it difficult for affected individuals to read, write, and spell, even when they possess average or above-average intelligence. Children with dyslexia often experience negative emotions such as low selfesteem, frustration, and anger, highlighting the importance of early detection and intervention. Various methods have been proposed for detecting dyslexia, including reading and writing tests, eye tracking, facial image analysis, and advanced imaging techniques such as Magnetic Resonance Imaging (MRI) and Electroencephalography (EEG). However, one key challenge in dyslexia detection is the reliance on small, unbalanced datasets, which can lead to poor generalization in models. This review suggests that deep learning techniques, particularly those utilizing advanced data augmentation methods, could address this issue by enhancing the model's ability to learn from limited data, improving both accuracy and robustness in real-world applications. The review focuses on the use of Machine Learning (ML) and Deep Learning (DL) for the automatic detection of dyslexia, aiming to improve diagnostic accuracy and accessibility. It explores various ML approaches, such as decision trees and support vector machines, as well as DL architectures, including Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs), which analyze text, speech, and eye-tracking patterns. The review provides a comparison of feature extraction methods, classification techniques, and performance metrics, with an emphasis on data enhancement and interpretability. Additionally, it addresses challenges such as dataset variability, ethical concerns, and the practical application of these methods in real-world settings. The goal of this review is to provide an overview of the most current methods for dyslexia detection and propose innovative ways to incorporate ML and DL into personalized learning plans an

关键词： Electroencephalography

来源：评论

学校读者我要写书评

暂无评论

Using Fundus Photography for Early Glaucoma Diagnosis by VGG-16 Architecture

Using Fundus Photography for Early Glaucoma Diagnosis by VGG...

引用

2024 International Conference on Advances in Modern Age technologies for Health and Engineering science, AMATHE 2024

作者： Ashwin, G. Reddy, Pathakottu Chenna Nagadevi, S. Srm Institute Of Science And Technology Department Of Computing Technologies Chennai India

ISBN: (纸本)9798350371567

Glaucoma stands as a prominent cause of irreversible vision impairment globally, marked by gradual harm to the optic nerve leading to visual field abnormalities. Timely identification and precise diagnosis play a pivotal role in ensuring effective treatment and the preservation of eyesight. The classification of the medical history of patients affected by glaucoma involves the utilization of a combination of machine learning algorithms, namely RandomForest and XGBoost, to identify individuals at risk. This research article aims to enhance the diagnostic capabilities of Glaucoma through the optimized utilization of contemporary methodologies, with a focus on leveraging deep learning techniques, specifically the VGG-16 model, in the analysis of Fundus Photograph captured by a panoptic direct ophthalmoscope. The novelty of this research lies in its integration of an individual's medical history, coupled with the detailed analysis of fundus photography, to enhance diagnostic precision and medical insights. In conclusion, the research is integrated with a mobile application for timely and accurate diagnosis of glaucoma which is essential in preventing further vision loss. © 2024 IEEE.

关键词： Eye protection

来源：评论

学校读者我要写书评

暂无评论

Density Vision: Crowd Classification using Deep Learning 3

Density Vision: Crowd Classification using Deep Learning

引用

3rd International Conference on Applied Artificial Intelligence and computing, ICAAIC 2024

作者： Singh, Harsh Vardhan Vidhya, S. Kumar, G Santosh Srm Institute of Science and Technology Department of Computing Technologies Kattankulathur India

ISBN: (纸本)9798350375190

Estimating crowd density and counting people are essential for crowd control, urban planning, and public safety. This research study utilizes a Multi-Column Convolutional Neural Network (MC-CNN) as a crowd counting technique trained on crowd datasets. The MC-CNN predicts crowd density maps from input images, providing precise crowd density estimates. Additionally, post-processing methods like clustering analyze the spatial crowd distribution. The proposed system incorporates object detection algorithms to generate individual coordinates, aiding in crowd analysis through clustering and then classifying them into three categories i.e. sparse, medium, and dense. Experimental findings showcase the method's efficacy in accurately estimating crowd densities and discerning crowd patterns. The proposed system provides valuable insights for crowd monitoring, resource allocation, and decision-making in densely populated areas. © 2024 IEEE.

关键词： Object detection

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：