CNN-Based Optimization for Fish Species Classification: Tackling Environmental Variability, Class Imbalance, and Real-Time Constraints DOI Creative Commons

Amirhosein Mohammadisabet,

Raza Hasan, Vishal Dattana

и другие.

Information, Год журнала: 2025, Номер 16(2), С. 154 - 154

Опубликована: Фев. 19, 2025

Automated fish species classification is essential for marine biodiversity monitoring, fisheries management, and ecological research. However, challenges such as environmental variability, class imbalance, computational demands hinder the development of robust models. This study investigates effectiveness convolutional neural network (CNN)-based models hybrid approaches to address these challenges. Eight CNN architectures, including DenseNet121, MobileNetV2, Xception, were compared alongside traditional classifiers like support vector machines (SVMs) random forest. DenseNet121 achieved highest accuracy (90.2%), leveraging its superior feature extraction generalization capabilities, while MobileNetV2 balanced (83.57%) with efficiency, processing images in 0.07 s, making it ideal real-time deployment. Advanced preprocessing techniques, data augmentation, turbidity simulation, transfer learning, employed enhance dataset robustness imbalance. Hybrid combining CNNs intermediate improved interpretability. Optimization pruning quantization, reduced model size by 73.7%, enabling deployment on resource-constrained devices. Grad-CAM visualizations further enhanced interpretability identifying key image regions influencing predictions. highlights potential CNN-based scalable, interpretable classification, offering actionable insights sustainable management conservation.

Язык: Английский

Machine learning and artificial intelligence in type 2 diabetes prediction: a comprehensive 33-year bibliometric and literature analysis DOI Creative Commons
Mahreen Kiran, Ying Xie, Nasreen Anjum

и другие.

Frontiers in Digital Health, Год журнала: 2025, Номер 7

Опубликована: Март 27, 2025

Background Type 2 Diabetes Mellitus (T2DM) remains a critical global health challenge, necessitating robust predictive models to enable early detection and personalized interventions. This study presents comprehensive bibliometric systematic review of 33 years (1991-2024) research on machine learning (ML) artificial intelligence (AI) applications in T2DM prediction. It highlights the growing complexity field identifies key trends, methodologies, gaps. Methods A methodology guided literature selection process, starting with keyword identification using Term Frequency-Inverse Document Frequency (TF-IDF) expert input. Based these refined keywords, was systematically selected PRISMA guidelines, resulting dataset 2,351 articles from Web Science Scopus databases. Bibliometric analysis performed entire tools such as VOSviewer Bibliometrix, enabling thematic clustering, co-citation analysis, network visualization. To assess most impactful literature, dual-criteria combining relevance impact scores applied. Articles were qualitatively assessed their alignment prediction four-point scale quantitatively evaluated based citation metrics normalized within subject, journal, publication year. scoring above predefined threshold for detailed review. The spans four time periods: 1991–2000, 2001–2010, 2011–2020, 2021–2024. Results findings reveal exponential growth publications since 2010, USA UK leading contributions, followed by emerging players like Singapore India. Key clusters include foundational ML techniques, epidemiological forecasting, modelling, clinical applications. Ensemble methods (e.g., Random Forest, Gradient Boosting) deep Convolutional Neural Networks) dominate recent advancements. Literature reveals that, studies primarily used demographic variables, while efforts integrate genetic, lifestyle, environmental predictors. Additionally, advances integrating real-world datasets, trends federated learning, explainability SHAP (SHapley Additive exPlanations) LIME (Local Interpretable Model-agnostic Explanations). Conclusion Future work should address gaps generalizability, interdisciplinary research, psychosocial integration, also focusing clinically actionable solutions applicability combat diabetes epidemic effectively.

Язык: Английский

Процитировано

0
CNN-Based Optimization for Fish Species Classification: Tackling Environmental Variability, Class Imbalance, and Real-Time Constraints DOI Creative Commons

Amirhosein Mohammadisabet,

Raza Hasan, Vishal Dattana

и другие.

Information, Год журнала: 2025, Номер 16(2), С. 154 - 154

Опубликована: Фев. 19, 2025

Automated fish species classification is essential for marine biodiversity monitoring, fisheries management, and ecological research. However, challenges such as environmental variability, class imbalance, computational demands hinder the development of robust models. This study investigates effectiveness convolutional neural network (CNN)-based models hybrid approaches to address these challenges. Eight CNN architectures, including DenseNet121, MobileNetV2, Xception, were compared alongside traditional classifiers like support vector machines (SVMs) random forest. DenseNet121 achieved highest accuracy (90.2%), leveraging its superior feature extraction generalization capabilities, while MobileNetV2 balanced (83.57%) with efficiency, processing images in 0.07 s, making it ideal real-time deployment. Advanced preprocessing techniques, data augmentation, turbidity simulation, transfer learning, employed enhance dataset robustness imbalance. Hybrid combining CNNs intermediate improved interpretability. Optimization pruning quantization, reduced model size by 73.7%, enabling deployment on resource-constrained devices. Grad-CAM visualizations further enhanced interpretability identifying key image regions influencing predictions. highlights potential CNN-based scalable, interpretable classification, offering actionable insights sustainable management conservation.

Язык: Английский

Процитировано

0