ABSTRACT This paper explores an AI‐assisted resource scheduling and cooperative learning model in a space–air–ground combined network (SAGIN) to possess ultra‐reliable low‐latency Internet of Medical Things (IoMT) applications. The generated healthcare data by the IoMT devices are processed by three levels in the considered scenario including the LEO satellites, the UAV swarms, and the ground MEC servers and adhere to strict latency, reliability, and privacy requirements. We aim at designing a multi‐tier resource allocation policy and federated learning policy that coordinates end‐to‐end latency and energy consumption and at the same time is highly accurate in terms of the model given privacy constraints. In this direction, we come up with a multi agent—deep deterministic policy gradient (MA‐DDPG) agent that allocates resources in a distributed manner and a hierarchical federated learning (HFL) system with delay‐sensitive aggregation to train models privately. Extensive simulation findings indicate that the presented framework can achieve 4.2 ms latency, 99.92% reliability, and 96.2% federated (global) model accuracy and 67% minimization of communication overhead, all of which are superior to baseline and the state‐of‐the‐art approaches in a variety of measures.
Tanvir Habib Sardar, Gousia Thahniyath, Ahlam Almusharraf et al.· Transactions on Emerging Tel...· 0 citations
Federated learning (FL) offers a promising approach to training machine learning models on decentralized data sources without directly exchanging data. However, FL systems are vulnerable to Byzantine attacks, where malicious participants can inject faulty model updates, compromising model accuracy and potentially causing significant harm. This paper proposes a novel framework for distributed federated learning with Byzantine fault tolerance, utilizing a blockchain-based verification layer. Our system employs a distributed consensus mechanism on a blockchain to validate model updates generated by participating nodes. This approach provides a robust, trustless environment, guaranteeing data integrity and ensuring model accuracy even in the presence of malicious actors. The core claim is the implementation of FL systems with Byzantine fault tolerance through blockchain verification. The core mechanism involves local model training followed by blockchain recording of updates and distributed consensus verification. This work represents a significant advancement by integrating the benefits of FL with the security and immutability offered by blockchain technology. We demonstrate a viable path toward building resilient and reliable FL systems for diverse applications.
Jincheng Zhang· Zenodo (CERN European Organi...· 0 citations
AI-driven failure detection is becoming essential in industrial manufacturing systems where conventional diagnostic methods often fall short in reliability and live feedback. This paper presents the integration of a modular artificial intelligence framework adapted to overcome these challenges by enabling intelligent fault monitoring systems. The proposed architecture consists of four integrated layers: a Sensor-Derived Adaptive Envelope Layer (SDAEL) for preprocessing and converting complex input signals; an Intelligent Deviation Mapping Engine (IDME), which utilizes a integration of One-Class Support Vector Machine (SVM) and Long Short-Term Memory (LSTM) networks for failure detection; a Failure Insight Notification Core (FINC) for real-time fault categorization; and a Federated Learning and Recovery Engine (FLARE) that enables secure, decentralized data updates across interlinked nodes. This framework can enables early fault detection in various manufacturing machine sectors. To validate this approach, an experimental deployment of the framework on a plastic bottle manufacturing machine that operated under changing conditions was conducted. The experimental evaluation showed improved detection accuracy and faster responsiveness compared to conventional diagnostic methods. Also Comparative benchmarking demonstrated that FLARE achieved a higher Precision of 0.80, Recall of 0.85, and an F1-score of 0.85, outperforming isolated LACSF models (F1-score 0.72) and LSTM-autoencoders (F1-score 0.62) by 13 and 23 points, respectively.This highlights the framework’s potential to improve fault diagnostics.
Faisal Ahamed Shaikh, SUDIPT PANTA, R. Ravi Kumar et al.· Scientific Reports· 0 citations
Federated learning (FL) offers a promising approach to training machine learning models on decentralized data, preserving data privacy. However, traditional differential privacy (DP) mechanisms often introduce significant noise into the global model updates, leading to substantial accuracy degradation. This paper presents a novel personalized differential privacy (P-DP) scheme for FL, where the noise level is dynamically adjusted based on the sensitivity of individual user data. We introduce a method for calculating per-user sensitivity values, considering local data distributions, and then employ adaptive noise scaling to minimize the privacy-utility trade-off. The proposed approach aims to achieve a better balance between privacy guarantees and model accuracy compared to standard DP methods in FL. The core contribution lies in the personalized allocation of the differential privacy budget, optimizing the system for a given application. Experimental results, although not presented here due to the focus on the method itself, would demonstrate the effectiveness of this approach.
Jincheng Zhang· Zenodo (CERN European Organi...· 0 citations
Reach audiences
Advertise in front of researchers, engineers, and readers.
Electroencephalography (EEG)-based Brain-Computer Interface (BCI) systems can support attention monitoring through non-invasive brain signal analysis.This study focuses on passive EEG attention classification for personalized cognitive tracking.A major problem is that EEG signals vary across clients, and non-independent and identically distributed data reduces the consistency of a single global model.Privacy is also important because raw EEG data should not be moved to a central server.Conventional centralized learning needs centralized data collection, and the performance of isolated local learning is weak, with accuracy of 0.8840 and macro F1 score of 0.8610.The global Federated Averaging model further improves these results to an accuracy of 0.9214 and macro F1 score of 0.9107, but with a gap in performance at the client level.This work proposes EdgeNeuroEEGNet, a one-dimensional Convolutional Neural Network (CNN) with attention, trained using federated learning and client-level personalization.The attention-state tracking is performed using a Cognitive Digital Twin layer.The data consists of 34 clients, 14 chosen EEG channels, 27.15 hours of recordings at 128 Hz and 12,512,552 EEG samples.The customized model had an accuracy of 0.9476, precision of 0.9526, recall of 0.9164 and macro F1 score of 0.9260.The results demonstrate that personalization increases accuracy, precision and macro F1, while simultaneously maintaining raw EEG data on the local level.
Vijayakumar Kempuraj, C. Lakshmi· International Journal of Int...· 0 citations
Reproducibility package for the manuscript "DEx-FCL: Drift- and Explanation-Consistent Federated Continual Learning for Adaptive Edge-IoT Security Management." This release includes the DEx-FCL source code, deterministic federated partitioning utilities, fixed five-seed configurations, CICIoT2023 and Edge-IIoTset benchmark result CSVs, generated figures, experiment launchers, dataset preparation instructions, and the deterministic Edge-IIoTset 10% sampling script. Raw benchmark datasets are not redistributed and must be obtained from their original providers.
Arul Selvam P· Zenodo (CERN European Organi...· 0 citations
This paper presents a novel approach to decentralized federated learning that leverages the strengths of differential privacy and homomorphic encryption to achieve robust privacy guarantees. The core idea is to encrypt each participant's local data using homomorphic encryption, allowing a central server to perform computations directly on the encrypted data without ever needing to access the plaintext. Subsequently, the encrypted results are sent back to the participants. This architecture effectively mitigates privacy risks associated with traditional federated learning methods. The combination of these three technologies – federated learning, differential privacy, and homomorphic encryption – provides a powerful framework for secure and decentralized machine learning. We demonstrate the potential for achieving strong privacy guarantees while maintaining the benefits of distributed learning.
Jincheng Zhang· Zenodo (CERN European Organi...· 0 citations
This paper explores the application of federated learning (FL) as a novel approach to privacy-preserving data analysis. Traditional data analysis methods often require centralized data collection, raising significant privacy concerns. Federated learning offers a compelling alternative by enabling collaborative model training without direct data sharing. The core claim of this work is the utilization of FL to conduct data analysis while safeguarding user privacy. The proposed mechanism involves constructing a FL framework where participants train models locally on their own datasets and subsequently aggregate model parameters. This process ensures that raw data remains decentralized, mitigating privacy risks. We delve into the technical aspects of FL, focusing on key considerations such as model aggregation techniques, communication efficiency, and privacy guarantees. The research contributes to a growing body of work in decentralized learning and provides a framework for addressing privacy challenges in various data-intensive applications. The primary goal is to demonstrate the feasibility and benefits of FL for secure and collaborative data analysis. ---
Jincheng Zhang· Zenodo (CERN European Organi...· 0 citations
Reproducibility package for the manuscript "DEx-FCL: Drift- and Explanation-Consistent Federated Continual Learning for Adaptive Edge-IoT Security Management." This release includes the DEx-FCL source code, deterministic federated partitioning utilities, fixed five-seed configurations, CICIoT2023 and Edge-IIoTset benchmark result CSVs, generated figures, experiment launchers, dataset preparation instructions, and the deterministic Edge-IIoTset 10% sampling script. Raw benchmark datasets are not redistributed and must be obtained from their original providers.
Arul Selvam P· Zenodo (CERN European Organi...· 0 citations
Federated learning (FL) offers a promising approach to training machine learning models on decentralized data, preserving data privacy. However, traditional differential privacy (DP) mechanisms often introduce significant noise into the global model updates, leading to substantial accuracy degradation. This paper presents a novel personalized differential privacy (P-DP) scheme for FL, where the noise level is dynamically adjusted based on the sensitivity of individual user data. We introduce a method for calculating per-user sensitivity values, considering local data distributions, and then employ adaptive noise scaling to minimize the privacy-utility trade-off. The proposed approach aims to achieve a better balance between privacy guarantees and model accuracy compared to standard DP methods in FL. The core contribution lies in the personalized allocation of the differential privacy budget, optimizing the system for a given application. Experimental results, although not presented here due to the focus on the method itself, would demonstrate the effectiveness of this approach.
Jincheng Zhang· Zenodo (CERN European Organi...· 0 citations
This paper presents a novel approach to generative modeling by integrating Generative Adversarial Networks (GANs) with Federated Learning (FL). Traditional GAN training suffers from centralized data requirements and privacy concerns. This research addresses these limitations through a distributed GAN architecture specifically designed for federated environments. The core idea is to train the generator and discriminator models concurrently across multiple devices (clients) without directly sharing their raw data. Instead, each client performs local GAN training and only shares model updates with a central server. This approach maintains data privacy while enabling the generation of high-quality synthetic data. The architecture utilizes a client-server framework where clients contribute to the global model through iterative updates. The proposed system aims to achieve superior performance compared to traditional GANs, particularly in scenarios with limited data and stringent privacy requirements. We explore the optimization strategies for the federated GAN training process, including addressing issues like non-IID data and model divergence. The theoretical framework and the proposed architecture are presented, outlining the key components and their interactions. Experimental results (simulated) demonstrate the feasibility and effectiveness of the approach in generating realistic synthetic data.
Jincheng Zhang· Zenodo (CERN European Organi...· 0 citations
Credit card transaction fraud has resulted in a massive loss to both consumers and banks in recent years.As a result, this research proposes an optimized framework for fraud detection.This framework will allow banks to construct fraud detection models using training data stored in their internal database.With this approach, financial institutions can collectively reap the benefits of a shared global model, which has seen more fraud than each bank alone, without sharing the dataset.Hence, the sensitive information of the cardholders is protected.The proposed optimization strategy focuses on decreasing communication costs when proceeding with federal training to accelerate convergence speed by optimizing the initial global model before the federated learning phase.Additionally, there is a significant degree of skewness in credit card data, which makes it difficult to predict fraudulent transactions.Unbalanced or skewed data is preprocessed using the resampling approach to obtain better results.This work uses seven meta-heuristic optimization algorithms.These algorithms' performance is documented along with a comparative analysis.The work is done in Python, and computation time, accuracy, precision, recall, F-measure, loss, and computation time are used to evaluate how well the algorithms perform.The experimental results show that the Heapbased Optimizer (HBO) with Federated Learning (FL) Model can achieve high detection performance and the minimum loss ratio across different datasets.The proposed HBO-FL framework achieves an average performance across three benchmark datasets of precision (0.9807), recall (0.9771), accuracy (0.9781), and F-score (0.9788).The results are on three publicly available datasets (European cardholders, BankSim, and Creditcardcsvpresent).For more reliability, the suggested approach is compared with the sex of the previous works.
Mustafa Abdul Salam, Doaa Elbably· International journal of int...· 0 citations
A new machine-learning framework aims to improve the success rate of computational protein design while moving away from results that reproduce sequences found in nature.
MIT News · Artificial Intelligence· news.mit.eduAug 24, 2026
A new method for surgically removing training examples from a model reveals that as datasets grow, the link between what a model learns and what it produces dissolves.
