BiGRU-CNN-AT: classifiying emotion on social media DOI
Rona Nisa Sofia Amriza,

Khairun Nisa Meiah Ngafidin

Data Technologies and Applications, Год журнала: 2024, Номер unknown

Опубликована: Дек. 30, 2024

Purpose This research aims to develop a robust deep-learning approach for classifying emotion in social media. Design/methodology/approach study integrates three deep learning techniques: Bidirectional Gated Recurrent Units (BiGRU), convolutional neural networks (CNN) and an attention mechanism, resulting the Convolution Attention (BiGRU-CNN-AT) model. The BiGRU captures potential semantic features, CNN extracts local features mechanism identifies keywords critical classification. Findings BiGRU-CNN-AT model outperformed several state-of-the-art classification algorithms. was compared against various baselines across multiple datasets, with methods consistently surpassing traditional approaches. Bi-LSTM demonstrated superior performance, particularly when combined mechanisms. Additionally, analysis of execution times indicated that processed data more efficiently. They were configuring hyperparameters integrating GloVe word embeddings, which significantly enhanced adam optimizer proving effective optimization. Originality/value paper contributes development novel framework, BiGRU-CNN-AT, bidirectional GRU, mechanisms text-based By leveraging strengths each component, this framework enhances accuracy tasks. Furthermore, offers comprehensive experimental analyses datasets.

Язык: Английский

Evidential-bio-inspired algorithms for modeling groundwater total hardness: A pioneering implementation of evidential neural network for feature selection in water resources management DOI Creative Commons
A. G. Usman, Abdulhayat M. Jibrin, Sagiru Mati

и другие.

Environmental Chemistry and Ecotoxicology, Год журнала: 2025, Номер unknown

Опубликована: Фев. 1, 2025

Язык: Английский

Процитировано

2
Prostate cancer grading framework based on deep transfer learning and Aquila optimizer DOI Creative Commons
Hossam Magdy Balaha,

Ahmed Osama Shaban,

Eman M. El-Gendy

и другие.

Neural Computing and Applications, Год журнала: 2024, Номер 36(14), С. 7877 - 7902

Опубликована: Фев. 22, 2024

Abstract Prostate cancer is the one of most dominant among males. It represents leading death causes worldwide. Due to current evolution artificial intelligence in medical imaging, deep learning has been successfully applied diseases diagnosis. However, recent studies prostate classification suffers from either low accuracy or lack data. Therefore, present work introduces a hybrid framework for early and accurate segmentation using learning. The proposed consists two stages, namely stage stage. In stage, 8 pretrained convolutional neural networks were fine-tuned Aquila optimizer used classify patients normal ones. If patient diagnosed with cancer, segmenting cancerous spot overall image U-Net can help diagnosis, here comes importance trained on 3 different datasets order generalize framework. best reported accuracies are 88.91% MobileNet “ISUP Grade-wise Cancer” dataset 100% ResNet152 “Transverse Plane Dataset” precisions 89.22% 100%, respectively. model gives an average AUC 98.46% 0.9778, respectively, “PANDA: Resized Train Data (512 × 512)” dataset. results give indicator acceptable performance

Язык: Английский

Процитировано

15
AutYOLO-ATT: an attention-based YOLOv8 algorithm for early autism diagnosis through facial expression recognition DOI Creative Commons

Reham Hosney,

Fatma M. Talaat, Eman M. El-Gendy

и другие.

Neural Computing and Applications, Год журнала: 2024, Номер 36(27), С. 17199 - 17219

Опубликована: Июнь 6, 2024

Abstract Autism Spectrum Disorder (ASD) is a developmental condition resulting from abnormalities in brain structure and function, which can manifest as communication social interaction difficulties. Conventional methods for diagnosing ASD may not be effective the early stages of disorder. Hence, diagnosis crucial to improving patient's overall health well-being. One alternative method autism facial expression recognition since autistic children typically exhibit distinct expressions that aid distinguishing them other children. This paper provides deep convolutional neural network (DCNN)-based real-time emotion system kids. The proposed designed identify six emotions, including surprise, delight, sadness, fear, joy, natural, assist medical professionals families recognizing intervention. In this study, an attention-based YOLOv8 (AutYOLO-ATT) algorithm proposed, enhances model's performance by integrating attention mechanism. outperforms all classifiers metrics, achieving precision 93.97%, recall 97.5%, F1-score 92.99%, accuracy 97.2%. These results highlight potential real-world applications, particularly fields where high essential.

Язык: Английский

Процитировано

6
Enhanced handwriting recognition through hybrid UNet-based architecture with global classical features DOI
Xiaofei Liu

Journal of Ambient Intelligence and Humanized Computing, Год журнала: 2025, Номер unknown

Опубликована: Янв. 18, 2025

Язык: Английский

Процитировано

0
Emotion-Assisted multi-modal Personality Recognition using adversarial Contrastive learning DOI
Yongtang Bao,

Yang Wang,

Yutong Qi

и другие.

Knowledge-Based Systems, Год журнала: 2025, Номер unknown, С. 113504 - 113504

Опубликована: Апрель 1, 2025

Язык: Английский

Процитировано

0
Artificial Intelligence-Based Suicide Prevention and Prediction: A Systematic Review (2019-2023) DOI

Anirudh Atmakuru,

Alen Shahini, Subrata Chakraborty

и другие.

Опубликована: Янв. 1, 2024

Язык: Английский

Процитировано

3
$$D_MD_RDF$$: diabetes mellitus and retinopathy detection framework using artificial intelligence and feature selection DOI Creative Commons
Hossam Magdy Balaha, Eman M. El-Gendy, Mahmoud M. Saafan

и другие.

Soft Computing, Год журнала: 2024, Номер 28(19), С. 11393 - 11420

Опубликована: Авг. 5, 2024

Abstract Diabetes mellitus is one of the most common diseases affecting patients different ages. can be controlled if diagnosed as early possible. One serious complications diabetes retina diabetic retinopathy. If not early, it lead to blindness. Our purpose propose a novel framework, named $$D_MD_RDF$$ DMDRDF , for and accurate diagnosis The framework consists two phases, detection (DMD) other retinopathy (DRD). novelty DMD phase concerned in contributions. Firstly, feature selection approach called Advanced Aquila Optimizer Feature Selection ( $$A^2OFS$$ xmlns:mml="http://www.w3.org/1998/Math/MathML">A2OFS ) introduced choose promising features diagnosing diabetes. This extracts required from results laboratory tests while ignoring useless features. Secondly, classification (CA) using five modified machine learning (ML) algorithms used. modification ML proposed automatically select parameters these Grid Search (GS) algorithm. DRD lies 7 CNNs reported concerning datasets shows that AO reports best performance metrics process with help classifiers. achieved accuracy 98.65% GS-ERTC model max-absolute scaling on “Early Stage Risk Prediction Dataset” dataset. Also, datasets, AOMobileNet considered suitable this problem outperforms CNN models 95.80% “The SUSTech-SYSU dataset”

Язык: Английский

Процитировано

1
Text classification algorithm of tourist attractions subcategories with modified TF-IDF and Word2Vec DOI Creative Commons
Lu Xiao,

Qiaoxing Li,

Qian Ma

и другие.

PLoS ONE, Год журнала: 2024, Номер 19(10), С. e0305095 - e0305095

Опубликована: Окт. 18, 2024

Text classification, as an important research area of text mining, can quickly and effectively extract valuable information to address the challenges organizing managing large-scale data in era big data. Currently, related on classification tends focus application fields such filtering, retrieval, public opinion monitoring, library information, with few studies applying methods field tourist attractions. In light this, a corpus attraction description texts is constructed using web crawler technology this paper. We propose novel representation method that combines Word2Vec word embeddings TF-IDF-CRF-POS weighting, optimizing traditional TF-IDF by incorporating total relative term frequency, category discriminability, part-of-speech information. Subsequently, proposed algorithm respectively seven commonly used classifiers (DT, SVM, LR, NB, MLP, RF, KNN), known for their good performance, achieve multi-class six subcategories national A-level The effectiveness superiority are validated comparing overall specific model stability against several methods. results demonstrate newly achieves higher accuracy F1-measure type professional dataset, even outperforms high-performance BERT currently favored industry. Acc, marco-F1, mirco-F1 values 2.29%, 5.55%, 2.90% higher. Moreover, identify rare categories imbalanced dataset exhibit better across datasets different sizes. Overall, presented paper exhibits superior performance robustness. addition, conclusions obtained predicted value true consistent, indicating practical. domain poses due its complexity (uneven length, relatively categories), high degree similarity between categories. However, efficiently implement multiple set, which beneficial exploration complex Chinese fields, provides useful reference vector expression similar content.

Язык: Английский

Процитировано

1
Deep attention for enhanced OCT image analysis in clinical retinal diagnosis DOI Creative Commons
Fatma M. Talaat,

Ahmed Abd Al-Rahman Ali,

Raghda Shawky El-Gendy

и другие.

Neural Computing and Applications, Год журнала: 2024, Номер unknown

Опубликована: Окт. 19, 2024

Abstract Retinal illnesses such as age-related macular degeneration (AMD) and diabetic maculopathy pose serious risks to vision in the developed world. The diagnosis assessment of these disorders have undergone revolutionary change with development optical coherence tomography (OCT). This study proposes a novel method for improving clinical precision retinal disease by utilizing strength Attention-Based DenseNet, deep learning architecture attention processes. For model building evaluation, dataset 84495 high-resolution OCT images divided into NORMAL, CNV, DME, DRUSEN classes was used. Data augmentation techniques were employed enhance model's robustness. DenseNet achieved validation accuracy 0.9167 batch size 32 50 training epochs. discovery presents promising route more precise speedy identification illnesses, ultimately enhancing patient care outcomes settings integrating cutting-edge technology powerful neural network architectures.

Язык: Английский

Процитировано

1
Winds of Power: Data Analysis for the Relationship between Wind Speed, Gust, and Power Output DOI Creative Commons
Samah A. Gamel,

Yara A. Sultan

Journal of Engineering Research - Egypt/Journal of Engineering Research, Год журнала: 2023, Номер 7(5), С. 189 - 194

Опубликована: Ноя. 1, 2023

Wind turbines are the most cost-effective and quickly evolving renewable energy technology. Benefits of this technology include no carbon emissions, resource conservation, job creation, flexible applications, modularity, fast installation, rural power grid improvement, potential for agricultural or industrial use.

Язык: Английский

Процитировано

2