Prediction of ailments using federated transfer learning and weight penalty-rational Tanh-RNN

Expert Systems with Applications, Journal Year: 2025, Volume and Issue: unknown, P. 127253 - 127253

Published: March 1, 2025

Language: Английский

---

Hybrid CNN-LSTM-GRU With Attention Mechanism for Efficient Cardio Vascular Disease Prediction in IoMT

Advances in medical technologies and clinical practice book series, Journal Year: 2025, Volume and Issue: unknown, P. 503 - 530

Published: Feb. 14, 2025

In this study, we develop a hybrid deep learning model for IoMT which is capable of delivering efficient predictive capability. The effectiveness was enhanced through feature selection pipeline using Pearson correlation, chi-square tests, and ExtraTreesClassifier ranking importance. By eliminating redundant attributes transforming categorical data with LabelEncoder, computational efficiency performance are enhanced. integrates CNN, LSTM, GRU layers, augmented by an attention mechanism. CNN component extracts spatial patterns from the input data, while LSTM layers capture temporal sequential dependencies. mechanism further enhances focusing on most relevant features, improving interpretability overall prediction accuracy. proposed demonstrates high level performance, achieving accuracy 98.9% curated dataset.

Language: Английский

---

Federated learning with multi‐cohort real‐world data for predicting the progression from mild cognitive impairment to Alzheimer's disease

Alzheimer s & Dementia, Journal Year: 2025, Volume and Issue: 21(4)

Published: April 1, 2025

Leveraging routinely collected electronic health records (EHRs) from multiple health-care institutions, this approach aims to assess the feasibility of using federated learning (FL) predict progression mild cognitive impairment (MCI) Alzheimer's disease (AD). We analyzed EHR data OneFlorida+ consortium, simulating six sites, and used a long short-term memory (LSTM) model with averaging (FedAvg) algorithm. A personalized FL was address between-site heterogeneity. Model performance assessed area under receiver operating characteristic curve (AUC) feature importance techniques. Of 44,899 MCI patients, 6391 progressed AD. models achieved 6% improvement in AUC compared local models. Key predictive features included body mass index, vitamin B12, blood pressure, others. showed promise predicting AD by integrating heterogeneous across institutions while preserving privacy. Despite limitations, it offers potential for future clinical applications. applied record institutions. improved prediction performance, increase identified key features, such as pressure. shows effectiveness handling heterogeneity sites ensuring Personalized pooled generally performed better than global

Language: Английский

---

Federated learning-based multimodal approach for early detection and personalized care in cardiac disease

Frontiers in Physiology, Journal Year: 2025, Volume and Issue: 16

Published: April 23, 2025

Heart disease remains a leading cause of mortality globally, and early detection is critical for effective treatment management. However, current diagnostic techniques often suffer from poor accuracy due to misintegration heterogeneous health data, limiting their clinical usefulness. To address this limitation, we propose privacy-preserving framework based on multimodal data analysis federated learning. Our approach integrates cardiac images, ECG signals, patient records, nutrition using an attention-based feature fusion model. preserve privacy ensure scalability, employ learning with locally trained Deep Neural Networks optimized Stochastic Gradient Descent (SGD-DNN). The fused vectors are input into the SGD-DNN classification. proposed demonstrates high in across multiple datasets: 97.76% Database 1, 98.43% 2, 99.12% 3. These results indicate robustness generalizability enables diagnosis personalized lifestyle recommendations while maintaining strict confidentiality. combination offers scalable, privacy-centric solution heart management, strong potential real-world implementation.

Language: Английский

---

Exploring privacy mechanisms and metrics in federated learning

Artificial Intelligence Review, Journal Year: 2025, Volume and Issue: 58(8)

Published: May 3, 2025

Language: Английский

---

Federated Deep Reinforcement Learning for Energy-Efficient Edge Computing Offloading and Resource Allocation in Industrial Internet

Applied Sciences, Journal Year: 2023, Volume and Issue: 13(11), P. 6708 - 6708

Published: May 31, 2023

Industrial Internet mobile edge computing (MEC) deploys servers near base stations to bring resources the of industrial networks meet energy-saving requirements terminal devices. This paper considers a wireless MEC system in an intelligent factory that has multiple and smart In this paper, device choice either offloading task whole or part server, performing it locally. Through combined optimization offload ratio, number subcarriers, transmission power, frequency, can achieve minimum total energy consumption. A resource allocation approach combines federated learning (FL) deep reinforcement (DRL) is suggested address problem. According simulation results, proposed algorithm displays fast convergence. Compared with baseline algorithms, significant advantages optimizing performance

Language: Английский

---

Multi-Objective artificial bee colony optimized hybrid deep belief network and XGBoost algorithm for heart disease prediction

Frontiers in Digital Health, Journal Year: 2023, Volume and Issue: 5

Published: Nov. 16, 2023

The global rise in heart disease necessitates precise prediction tools to assess individual risk levels. This paper introduces a novel Multi-Objective Artificial Bee Colony Optimized Hybrid Deep Belief Network and XGBoost (HDBN-XG) algorithm, enhancing coronary accuracy. Key physiological data, including Electrocardiogram (ECG) readings blood volume measurements, are analyzed. HDBN-XG algorithm assesses data quality, normalizes using z-score values, extracts features via the Computational Rough Set method, constructs feature subsets approach. Our findings indicate that achieves an accuracy of 99%, precision 95%, specificity 98%, sensitivity 97%, F1-measure 96%, outperforming existing classifiers. contributes predictive analytics by offering data-driven approach healthcare, providing insights mitigate impact disease.

Language: Английский

---

FedCVD: Towards a Scalable, Privacy-Preserving Federated Learning Model for Cardiovascular Diseases Prediction

Published: Jan. 26, 2024

This paper presents FedCVD, a federated learning model designed for predicting cardiovascular disease (CVD) by employing logistic regression and Support Vector Machine (SVM) algorithms. FedCVD utilizes the privacy scalability advantages offered to facilitate collaborative training using decentralized patient data, ensuring confidentiality. To evaluate effectiveness of proposed model, experiments were conducted 10-year risk coronary heart Kaggle dataset. address data imbalance challenges, three techniques—Random Over Sampling, Random Under Synthetic Minority Oversampling Technique (SMOTE)—were explored. The study demonstrates promising performance,For with SMOTE achieving an AUC value 0.7048. Comparative analysis centralized shows competitive results, 0.7081 Sampling. For SVM 0.7340 is achieved In comparison, machine approach utilizing Sampling achieves 0.6962. These findings highlight approach, surpassing performance models CVD prediction.

Language: Английский

---

Federated Multi-Label Learning (FMLL): Innovative Method for Classification Tasks in Animal Science

Animals, Journal Year: 2024, Volume and Issue: 14(14), P. 2021 - 2021

Published: July 9, 2024

Federated learning is a collaborative machine paradigm where multiple parties jointly train predictive model while keeping their data. On the other hand, multi-label deals with classification tasks instances may simultaneously belong to classes. This study introduces concept of Multi-Label Learning (FMLL), combining these two important approaches. The proposed approach leverages federated principles address tasks. Specifically, it adopts Binary Relevance (BR) strategy handle nature data and employs Reduced-Error Pruning Tree (REPTree) as base classifier. effectiveness FMLL method was demonstrated by experiments carried out on three diverse datasets within context animal science: Amphibians, Anuran-Calls-(MFCCs), HackerEarth-Adopt-A-Buddy. accuracy rates achieved across were 73.24%, 94.50%, 86.12%, respectively. Compared state-of-the-art methods, exhibited remarkable improvements (above 10%) in average accuracy, precision, recall, F-score metrics.

Language: Английский

---

Fed-CL- an atrial fibrillation prediction system using ECG signals employing federated learning mechanism

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: Sept. 9, 2024

Language: Английский

---