检索结果-内蒙古大学图书馆

A Modified Deep Residual-Convolutional Neural Network for Accurate Imputation of Missing Data

computers, Materials & Continua 2025年第2期82卷 3419-3441页

作者： Firdaus Firdaus Siti Nurmaini Anggun Islami Annisa Darmawahyuni Ade Iriani Sapitri Muhammad Naufal Rachmatullah Bambang Tutuko Akhiar Wista Arum Muhammad Irfan Karim Yultrien Yultrien Ramadhana Noor Salassa Wandya Intelligent System Research Group Faculty of Computer ScienceUniversitas SriwijayaPalembang30139Indonesia

Handling missing data accurately is critical in clinical research, where data quality directly impacts decision-making and patient outcomes. While deep learning (DL) techniques for data imputation have gained attention, challenges remain, especially when dealing with diverse data types. In this study, we introduce a novel data imputation method based on a modified convolutional neural network, specifically, a Deep Residual-Convolutional Neural Network (DRes-CNN) architecture designed to handle missing values across various datasets. Our approach demonstrates substantial improvements over existing imputation techniques by leveraging residual connections and optimized convolutional layers to capture complex data patterns. We evaluated the model on publicly available datasets, including Medical Information Mart for Intensive Care (MIMIC-III and MIMIC-IV), which contain critical care patient data, and the Beijing Multi-Site Air Quality dataset, which measures environmental air quality. The proposed DRes-CNN method achieved a root mean square error (RMSE) of 0.00006, highlighting its high accuracy and robustness. We also compared with Low Light-Convolutional Neural Network (LL-CNN) and U-Net methods, which had RMSE values of 0.00075 and 0.00073, respectively. This represented an improvement of approximately 92% over LL-CNN and 91% over U-Net. The results showed that this DRes-CNN-based imputation method outperforms current state-of-the-art models. These results established DRes-CNN as a reliable solution for addressing missing data.

关键词： Data imputation missing data deep learning deep residual convolutional neural network

来源：评论

学校读者我要写书评

暂无评论

Transfer function adaptation for effective feature selection with the side-blotched lizard algorithm

引用

Multimedia Tools and Applications 2024年 1-21页

作者： Eisa, Mohamed Abdel-lah, Marwa Farouk, Fifi Abdel-aal, Amr System and Information Technology Department Faculty of Management Technology and Information System Port Said University Port Said Egypt Department of Computer Science Faculty of Computer and Informatics Zagazig University Zagazig Egypt

Feature selection is a crucial preprocessing step in data mining and machine learning, enhancing model performance and computational efficiency. This paper investigates the effectiveness of the Side-Blotched Lizard Optimization Algorithm (SBLA) for feature selection by developing six novel variants: Sbla-s1, Sbla-s2, Sbla-s3, Sbla-v1, Sbla-v2, and Sbla-v3, each employing distinct S-shaped or V-shaped transfer functions to convert the continuous search space to a binary format. These variants were rigorously evaluated on nineteen benchmark datasets from the UCI repository, comparing their performance based on average classification accuracy, average number of selected features, and average fitness value. The results demonstrated the superiority of Sbla-s3, achieving an average classification accuracy of 92.8% across all datasets, a mean number of selected features of 20, and an average fitness value of 0.08. Furthermore, Sbla-s3 consistently outperformed six other state-of-the-art metaheuristic algorithms, achieving the highest average accuracy on sixteen out of nineteen datasets. These findings establish Sbla-s3 as a promising and effective approach for feature selection, capable of identifying relevant features while maintaining high classification accuracy, potentially leading to improved model performance in various machine learning applications. © The Author(s) 2024.

关键词： Heuristic algorithms

来源：评论

学校读者我要写书评

暂无评论

Hybrid Approach of LDA, ROS-NCL, and CNN for Topic Mining in Fintech User Reviews

Hybrid Approach of LDA, ROS-NCL, and CNN for Topic Mining in...

引用

2025 International Conference on Advancement in Data Science, E-learning and Information system, ICADEIS 2025

作者： Hartono, Pieter Chang Widiantoro, Albertus Dwiyoga Harnadi, Bernadinus Nugroho, Erdhi Widyarto Faculty of Computer Science Semarang Information System Department Indonesia

ISBN: (纸本)9798331513320

The development of digital payment technology, such as Gopay, has provided convenience for the Indonesian public. To understand user satisfaction and identify key issues, this study analyzes Gopay user reviews using a hybrid approach. The main issues addressed are the imbalance in review data and the lack of studies integrating Latent Dirichlet Allocation (LDA) and Convolutional Neural Network (CNN) models for topic analysis. This research encompasses three main stages: topic identification using LDA, data balancing with Random Over Sampling and Neighborhood Cleaning Rule (ROS-NCL), and topic classification using CNN. The model is designed to accurately identify and classify key topics. Furthermore, this method is compared with other approaches to assess its advantages. The research findings show that this hybrid approach is effective in addressing data imbalance, achieving a topic classification accuracy of 96%. These findings provide valuable insights for service providers to improve the quality of Gopay services based on topic analysis derived from user feedback. © 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

Application of Artificial Intelligence Methods for Generating Mathematical Problems in Primary Education: Analysis of Effectiveness and Prospects for Automated Assessment 7

Application of Artificial Intelligence Methods for Generatin...

引用

7th International Conference on Machine Learning and Natural Language Processing, MLNLP 2024

作者： Salavatovich, Nazipov Rustam Yurievna, Avksentieva Elena ITMO University Faculty of Software Engineering and Computer System Saint Petersburg Russia ITMO University Faculty of Software Engineering and Computer System Saint-Petersburg Russia

ISBN: (纸本)9798350354973

The present study aims to evaluate the effectiveness of using modern neural network models (ChatGPT 4o, Claude 3.5 Sonnet, GigaChat 3.0, YandexGPT 3) for generating mathematical problems for elementary school students. The research conducts a comparative analysis of the quality of generated problems, assesses the possibility of their direct use without expert review, and develops recommendations for automating the quality control process. The methodology includes the generation of various types of mathematical problems, their expert evaluation according to developed criteria, and statistical analysis of the results. The study identifies the strengths and weaknesses of each model, determines the types of problems in which the models demonstrate the best results, and suggests directions for further improvement of AI systems in the educational context. © 2024 IEEE.

关键词： Neural network models

来源：评论

学校读者我要写书评

暂无评论

3D Object Detection for Autonomous Driving: A Comprehensive Review 6

3D Object Detection for Autonomous Driving: A Comprehensive ...

引用

6th International Conference on Computing and Informatics, ICCI 2024

作者： Nagiub, Ahmed Soliman Ghoniemy, Said Fayez, Mahmoud Khaled, Heba Computer System Department Faculty of Computer and information Science Ain Shams University Cairo Egypt

ISBN: (纸本)9798350373875

accurate perception is a critical component in enabling autonomous vehicles to understand their driving environment. The acquisition of 3D information about objects, including their location and pose, is essential for achieving this understanding. This survey paper presents a comprehensive review of 3D object detection techniques specifically tailored for autonomous vehicles. The survey begins with an introduction to 3D object detection, elucidating the significance of the third dimension in perceiving the driving environment. It explores the types of sensors utilized in this context and the corresponding data extracted from these sensors. Additionally, the survey investigates the different types of datasets employed, including their formats, sizes, and provides a comparative analysis. Furthermore, the paper categorizes and thoroughly examines the perception methods employed for 3D object detection, based on the diverse range of sensors utilized. Each method is evaluated based on its effectiveness in accurately detecting objects in a three-dimensional space. Additionally, the evaluation metrics used to assess the performance of these methods are discussed. By offering a comprehensive overview of 3D object detection techniques for autonomous vehicles, this survey aims to advance the field of perception systems. It serves as a valuable resource for researchers and practitioners, providing insights into the techniques, sensors, and evaluation metrics employed in 3D object detection for autonomous vehicles. © 2024 IEEE.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

A hybrid deep learning framework for solar irradiation prediction based on regional satellite images and data

引用

Neural Computing and Applications 2025年 1-37页

作者： Attya, Mohammed Abo-Seida, Osama Mohamed, Hatem Mohammed, Amgad Department of Information System Faculty of Computers and Information Kafrelsheikh University Kafrelsheikh Egypt Department of Computer Science Faculty of Computers and Information Kafrelsheikh University Kafrelsheikh Egypt

Accurate prediction of solar irradiation, particularly Diffuse Horizontal Irradiance (DHI), is essential for optimizing solar energy systems. This paper presents a hybrid framework that integrates satellite imagery and satellite-derived data to improve the precision of DHI forecasts. Two key datasets were employed: ground-based Solar Irradiation Measurement (SRM) and satellite-based Solar Irradiation (SSR). The proposed methodology utilizes machine learning models along two paths. The first path processes satellite imagery using advanced techniques, including pixel imputation with a modified Random Forest (RF) and Generative Adversarial Network (GAN), noisy region detection with Self-Organizing Maps (SOM), and noise removal through a Latent Diffusion Model (LDM). The second path handles tabular data through a diffusion probabilistic model designed for missing data imputation. These features are then merged and fed into a Long Short-Term Memory (LSTM) network, enhanced to capture seasonality and trends for accurate DHI predictions. The paper makes significant contributions by introducing a robust hybrid framework that leverages both satellite imagery and tabular data, incorporating novel preprocessing methods such as GAN-based pixel imputation and diffusion-based noise removal. The LSTM model is further adapted to handle seasonal and trend components, resulting in enhanced DHI forecasting accuracy. The proposed model demonstrates superior performance, achieving a Mean Squared Error (MSE) of 8.170 W/m2, Root Mean Squared Error (RMSE) of 1.749 W/m2, Mean Absolute Error (MAE) of 0.693 W/m2, and an R-squared value of 2.574, significantly outperforming conventional methods such as Artificial Neural Networks (ANN), Support Vector Regression (SVR), and traditional LSTM models. Additionally, the model records a low Mean Absolute Percentage Error (MAPE) of 0.8183% and a computation time of 35.4 s, highlighting its efficiency. These results contribute to more accurate and reli

关键词： Landsat

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

A Framework for Food Availability Forecasting Using Temporal Knowledge Graph

A Framework for Food Availability Forecasting Using Temporal...

引用

2024 International Conference on computer and Applications, ICCA 2024

作者： Eltaki, Nada Nagaty, Khaled Elzanfaly, Doaa Maghawry, Noura The British University in Egypt Faculty of Informatics and Computer Science Department of Information System Cairo Egypt The British University in Egypt Faculty of Informatics and Computer Science Department of Computer Science Cairo Egypt Helwan University Faculty of Informatics and Computer Science Department of Information System Cairo Egypt

ISBN: (纸本)9798350367560

The entire world encounters a significant problem which is food availability. 'Zero-hunger' is the second Sustainable Development Goal (SDG) of the goals established by the United Nations in 2015. With the increasing number of people and various factors affecting food availability such as climate and country-based food production rate, food forecasting models have become increasingly essential. However, traditional forecasting techniques are limited in dealing with the data and identifying the relationships between them within a specific timestamp. This paper proposes a framework for an intelligent food availability forecasting model built using deep learning techniques and based on a temporal knowledge graph where different timestamps are taken into consideration. The proposed framework helps in enhancing and overcoming the challenges and limitations of the recent forecasting techniques leading to a dynamic, scalable, intelligent forecasting model. © 2024 IEEE.

关键词： Prediction models

来源：评论

学校读者我要写书评

暂无评论

Facial Expression Classification Using Extreme Learning Machine and Genetic Algorithm 11

Facial Expression Classification Using Extreme Learning Mach...

引用

11th International Conference on Electrical Engineering, computer Science and Informatics, EECSI 2024

作者： Bachtiar, Fitra Abdurrachman Nuha, Mufidatun Universitas Brawijaya Intelligent System Laboratory Computer Science Faculty Malang Indonesia

ISBN: (纸本)9798350355314

Feature extraction is a critical component of facial expression recognition, requiring careful selection to ensure optimal performance. Traditional template-based models often struggle with computational efficiency and are sensitive to factors such as illumination, color variation, and image quality. Although geometry-based models have demonstrated robustness in feature extraction, further optimization remains needed to enhance the recognition accuracy. In this study, the CK+ dataset is utilized, and employ facial landmarks as a geometry-based feature extraction method. The extracted features are then processed using an Extreme Learning Machine combined with Genetic Algorithms (ELM-GA) to select a subset of features and classify facial expressions. The proposed model effectively reduces the number of features used in the classification model by more than half, improving computational efficiency. Additionally, the accuracy of the proposed model outperforms the VGGNet and combination of machine learning and Local Binary Patterns (LBP) and machine learning and Gray-Level Co-occurrence Matrix (GLCM) for feature extraction. Although the current model does not represent the best possible solution, the promising accuracy results provide a solid foundation for further refinement and development in facial expression recognition systems. © 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

MV-DUO: multi-variate discrete unified optimization for psychological vital assessments

引用

Neural Computing and Applications 2024年第31期36卷 19777-19793页

作者： Pethuraj, Mohamed Shakeel Burhanuddin, M.A. Dzakiyullah, Nur Rachman Department of Information System Faculty of Computer and Engineering Universitas Alma Ata Yogyakarta Indonesia Faculty of Information and Communication Technology Universiti Teknikal Malaysia Melaka Durian Tunggal Melaka Malaysia

Psychological vital assessments are required for monitoring health conditions and observing body reactions toward diseases and medications. Wearable sensors play a vital role in sensing body vitals and presenting them as signals for computer-based analysis. The problem relies on the signal decoding due to its input stream that turns out to be discrete/ continuous. Therefore for addressing the above specific issue, this article introduces MV-DUO (Multi-Variate Discrete Unified Optimization) method. This method addresses the above problem from a multi-variate perspective by sensing differential signals across healthy and unhealthy conditions. The healthy and unhealthy conditions are trained using neural learning by augmenting/ ceasing external vital data. The unification is performed using single-point artificial ecosystem-based optimization for identifying discrete sequences collaborated with continuous signals. The single-point reference is grouped based on the maximum continuity fitness observed under various sensing intervals. In this process, the non-grouped sequences are identified as unhealthy or discrete for which additional detection training and classification are required. Considerably the changes between successive sensing intervals are used for variations detection from unified high-fitness groups. Those grouped instances are used for training new vital changes observed at distinct intervals. This improves detection accuracy under controlled errors. For the varying sensing intervals, the proposed method achieves 14.13% high accuracy, 8.29% high grouping rate, 10.77% less error, and 10.07% less detection time. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2024.

关键词： computer aided analysis

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：