Privacy-Preserving Machine Learning Using Federated Learning and Secure Aggregation

Mobile and Web-of-Things (WoT) devices at the network edge account for more than half of the world's web traffic, making a great data source for various machine learning (ML) applications, particularly federated learning (FL) which offers a promising solution to privacy-preserving ML feeding on these data. FL allows edge mobile and WoT devices to train a shared global ML model under the orchestration of a central parameter server. In the real world, due to resource heterogeneity, these edge devices often train different versions of models (e.g., VGG-16 and VGG-19) or different ML models (e.g., VGG and ResNet) for the same ML task (e.g., computer vision and speech recognition). Existing FL schemes have assumed that participating edge devices share a common model architecture, and thus cannot facilitate FL across edge devices with heterogeneous ML model architectures. We explored this architecture heterogeneity challenge and found that FL can and should accommodate these edge devices to improve model accuracy and accelerate model training. This paper presents our findings and FlexiFed, a novel scheme for FL across edge devices with heterogeneous model architectures, and three model aggregation strategies for accommodating architecture heterogeneity under FlexiFed. Experiments with four widely-used ML models on four public datasets demonstrate 1) the usefulness of FlexiFed; and 2) that compared with the state-of-the-art FL scheme, FlexiFed improves model accuracy by 2.6%-9.7% and accelerates model convergence by 1.24 × -4.04 ×.

To address privacy concerns, federated learning (FL) is becoming a promising machine learning technique which enables multiple decentralized clients to train a shared model collaboratively while preserving their private training data. Although FL may reduce the risks of data leak, it is still possible for hackers to reverse-engineer a trained model and figure out the information in the original training dataset provided by a FL client. In order to avoid such risks, secure aggregation (SA) can be used to privately combine the trained models of the clients to update the shared model. However, SA usually introduces performance overhead as it requires additional computation for encryption operations and even communications when secure multi-party computation (SMPC) is used. In this paper, we analyze the performance of FL with SA using PySyft, an open source framework including FL implementation, and propose an asynchronous FL mechanism to improve the overall performance. It turns out that the performance depends on the computational capabilities of the clients and the characteristics of the communication network, and we propose a performance modeling method to help system designers break down the execution time and decide on suitable trade-offs between privacy, efficiency, and accuracy for a balanced system.

As federated learning (FL) keeps dominating the way in privacy-preserving machine learning (ML), notably in finance, its long-term sustainability is confronted with the approaching challenge of quantum computing. The introduction of quantum computing poses a challenge to traditional cryptographic technique on which FL relies RSA, elliptic curve cryptography (ECC), and some homomorphic encryption (HE) methods applied for encrypting model updates during training rendering this chapter to explore the interplay between the threat of quantum computing and the FL architectures, with emphasis on the implications this interaction introduces to cybersecurity and privacy in finance applications. Quantum computers are theoretically capable of solving hard math problems that have been regarded as computationally intractable for quantum computers. The well-known algorithms like Shor’s and Grover’s provides serious risk to the cryptographic procedures FL pipelines depends on. The finance sector, by its nature involving sensitive data and dependence on secure, distributed ML systems, is particularly vulnerable. The chapter begins by summarizing the fundamentals of FL and current encryption methods adapted to protect FL workflows. It then shifts towards the essence of quantum computing, emphasizing on the notable algorithms that threaten FL. We also explore potential attack points through which quantum computing could attack FL, e.g., model inversion attacks, poisoning attacks facilitated by quantum capabilities, and threats to secure aggregation. Post-quantum cryptography (PQC) is also presented as a potential counterattack, followed by a discussion of its use case in FL platforms. This chapter attempts to close the gap between FL and quantum computing threats to the financial sector. It is an introduction to practitioners and researchers to revisit examining security assumptions in FL and move towards the quantum-augmented cyber threat model.

Federated learning (FL) enables decentralized, privacy-preserving machine learning by training models across distributed data without sharing raw patient information. However, most FL frameworks focus on unimodal data and overlook critical challenges in multimodal healthcare settings, such as privacy risks, fairness disparities, and reduced model interpretability. We present PRIFLEX (Privacy-Resilient Integration Framework for Learning Exchange), a novel FL framework designed for secure integration of structured and unstructured medical data. PRIFLEX combines 12-lead electrocardiograms (ECG) from the PhysioNet PTB-XL dataset and clinical notes from the Medical Information Mart for Intensive Care IV (MIMIC-IV), supporting early, late, and hybrid data fusion strategies. To safeguard model updates, it evaluates standalone and hybrid defenses using Differential Privacy (DP) and Secure Aggregation (SA) against gradient leakage, model inversion, and membership inference attacks. Results show that early fusion improves the area under the Receiver Operating Characteristic curve (AUROC) by up to 6.2%, while hybrid DP+SA reduces attack success rates by up to 84% and improves fairness with manageable system overhead. PRIFLEX also quantifies interpretability loss using SHapley Additive exPlanations (SHAP) and gradient-based methods, highlighting the trade-off between privacy and transparency. Overall, PRIFLEX sets a new benchmark for building secure, fair, and explainable federated learning systems in healthcare.

This article provides a comprehensive survey of aggregation strategies in federated learning (FL). This decentralized machine learning (ML) paradigm enables multiple clients to collaboratively train models without sharing their local datasets. Aggregation is a pivotal aspect of FL, as it integrates model updates from diverse clients into a unified global model while addressing critical challenges such as data heterogeneity, scalability, and privacy preservation. The study categorizes aggregation strategies into three primary approaches: data-centric, model-centric, and secure aggregation, each tailored to address distinct problems in FL systems. Data-centric strategies focus on addressing non-independent and identically distributed (non-IID) data across clients, ensuring that model updates account for imbalances in data distributions. Model-centric strategies optimize the aggregation of model parameters, emphasizing techniques such as weighted averaging and model distillation to improve model performance across clients with diverse data characteristics. Secure aggregation techniques aim to enhance privacy and robustness, protecting client data from potential adversarial threats through encryption-based methods like secure multi-party computation (SMPC) and Byzantine-resilient techniques. The analysis delves into the advantages and limitations of these aggregation strategies, particularly their role in tackling challenges like non-IID data, communication efficiency, and resistance to adversarial attacks. Furthermore, the article identifies existing research gaps in FL, including the need for more scalable and robust aggregation methods capable of reducing communication costs, enhancing privacy guarantees, and improving performance in highly heterogeneous environments. These insights provide a roadmap for future research aimed at advancing aggregation strategies in FL to improve model accuracy, security, and efficiency in real-world applications.

This systematic literature review uses the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) methodology to assess progress in blockchain-based Federated learning (FL) and Machine Learning (ML) for detecting financial fraud over the last five years (2020–2025). An initial pool of 29,274 records identified across IEEE Xplore, ACM Digital Library, and ScienceDirect yielded 1585 peer-reviewed studies that met the inclusion criteria. Both qualitative and quantitative approaches were used. The examined papers were classified according to algorithm type, fraud types, and evaluation measures. Credit card fraud and cryptocurrency fraud dominated the literature, with supervised learning (e.g., XGBoost, 95% accuracy) and federated learning (e.g., FedAvg, 91% accuracy) emerging as dominant methodologies. Centralized ML outperforms FL in latency but poses privacy risks. FL–blockchain hybrids reduce false positives. While precision, recall, and F1-score are commonly used, few studies use cost-sensitive criteria. Future research should prioritize adaptive FL aggregation, privacy-preserving ML, and cross-industry collaboration.

Federated Learning (FL) is a transformative paradigm that enables collaborative machine learning across multiple decentralized data sources without requiring direct data sharing. This approach has emerged as a critical solution for privacy-preserving machine learning, particularly in sensitive domains such as healthcare, finance, and IoT systems. FL frameworks employ advanced techniques such as secure aggregation, differential privacy, and homomorphic encryption to protect user data while facilitating model convergence. These frameworks address key challenges, including communication efficiency, data heterogeneity, and robustness against malicious actors. This paper reviews the state-of-the-art FL frameworks, exploring their architecture, privacy-preserving techniques, and application scenarios. Furthermore, it highlights recent advancements in secure federated learning protocols, multi-party computation strategies, and scalability improvements. The discussion extends to open challenges, such as ensuring fairness, optimizing resource consumption, and maintaining security guarantees in adversarial settings. By advancing the development of privacy-preserving FL frameworks, researchers and practitioners can unlock the potential of collaborative machine learning while upholding stringent data privacy standards.

With the widespread application of Artificial Intelligence (AI) and Machine Learning (ML) in medical field, early diagnosis of brain tumors has become increasingly significant. However, traditional methods face challenges such as data privacy, model interpretability, and data heterogeneity. This paper presents a detailed literature review of the application of Federated Learning (FL) in brain tumor classification, focusing on the necessity of privacy-preserving ML using Magnetic Resonance Imaging (MRI) technology. This paper analyzes various Convolutional Neural Network (CNN) models, including VGG16, ResNet50, DenseNet121, and EfficientNet, exploring their integration within the FL framework to enhance diagnostic accuracy while preserving patient data privacy. Through the discussion, core issues in this field emerge, including model interpretability, non-independent and identically distributed (Non-IID) data distributions, and computational challenges in FL architecture. Although these factors limit the widespread application of FL in medical settings, this paper also proposes potential solutions, such as improving algorithm interpretability through interpretable tools, and utilizing transfer learning and domain adaptation methods to enhance model effectiveness across different datasets. Techniques like knowledge distillation, model quantization, and pruning are proposed to enhance computational efficiency and minimize communication costs. Future research should focus on the application of these methods to enhance model reliability and efficiency.

Federated learning (FL) is gradually becoming a key learning paradigm in Privacy-preserving Machine Learning (ML) systems. In FL, a large number of clients cooperate with a central server to learn a shared model without sharing their own data sets. However, since there is a great disparity between the client data sets, standard FL is often hard to tune and suffers from performance degradation due to the inharmony among local models. To this end, in this paper we propose a novel FL scheme, termed client reputation federated learning (CRFL), which dynamically assesses the reputation of the clients participating in FL. Our method leverages techniques from model explanation, and aims at precisely measure each client's impact to the global model. To be specific, we first calculate the saliency-weighted variance on pixelwise relevance scores as the quality factor of a single sample. Then we extract activation function values at the last hidden layer to compute the divergence factor of individual data set. Finally, the server integrates these two factors as an assessment of the client reputation. By leveraging such assessment, CRFL can dynamically adjust the weights of the clients in each aggregation round, thus leading to a significant improvement over the baseline method in terms of model accuracy and convergence rate. Intensive experiments are conducted on the MNIST and CIFAR-10 data sets, and experimental results demonstrate the efficacy of the proposed method.

In international nuclear safeguards, the International Atomic Energy Agency (IAEA) is tasked with inspecting and verifying nuclear facilities and their activities. Data analytics and machine learning to support inspections require large amounts of data that nuclear facility operators may consider proprietary or sensitive, so the IAEA may not have full access. Allowing computation over private data without compromising its security therefore has value for safeguards inspections and analysis. Privacy-preserving machine learning (PPML) consists of security-focused techniques that allow data analytics and machine learning algorithms to run on sensitive data without revealing it. This includes ideas like homomorphic encryption (HE), secure multiparty computation (SMPC), and secure enclaves. HE allows algorithms and mathematical operations to be conducted directly on the encrypted data instead of first decrypting it. With SMPC, multiple entities collaboratively compute over distributed data such that no party is able to directly view any others’ original data. Secure enclaves allow computation to take place in a separate and heavily blocked-off section of a CPU. Techniques like these allow for several potential use cases in which the security of data is essential. With SMPC, machine learning models can be trained over the input data from multiple entities, resulting in a model that all users can benefit from without leaking the input data from any particular entity. With SMPC or a zero-knowledge proof (ZKP), an algorithm returning some single answer or truth value can be run on someone else’s data without ever needing to see that data, potentially allowing for verification or proof of some underlying question. HE can allow for outsourcing computation on data to a hostile or untrusted environment. Although most of the research in this field resides within the health and financial domains, tools from PPML may have similar applications in nuclear safeguards. Allowing the IAEA to compute over proprietary information, such as process models and raw sensor data using PPML techniques, provides the baseline for running complex analytics without needing direct unencrypted access to the underlying data, maintaining its privacy. Important limitations to consider for these techniques include the efficiency and level of security required. The security of HE and SMPC come at the cost of speed—the significant amount of overhead means that algorithms implemented in these protocols and encryption schemes are slower than when run on plaintext. Additionally, several important parameters determine what techniques or protocols are used based on the security requirements. SMPC protocols may need to be selected for resistance against a party that attempts to deviate from the protocol to distort the result or gain access to additional information, and a protocol secure against these attacks may further increase the overhead of the algorithm.

Machine learning techniques have long been the foundations of speech processing. Bayesian classiflcation, decision trees, unsupervised clustering, the EM algorithm, maximum entropy, etc. are all part of existing speech recognition systems. The success of statistical speech recognition has led to the rise of statistical and empirical methods in natural language processing. Indeed, many of the machine learning techniques used in language processing, from statistical part-of-speech tagging to the noisy channel model for machine translation have roots in work conducted in the speech fleld. However, advances in learning theory and algorithmic machine learning approaches in recent years have led to signiflcant changes in the direction and emphasis of the statistical and learning centered research in natural language processing and made a mark on natural language and speech processing. Approaches such as memory based learning, a range of linear classiflers such as Boosting, SVMs and SNoW and others have been successfully applied to a broad range of natural language problems, and these now inspire new research in speech retrieval and recognition. We have seen an increasingly close collaboration between voice and language processing researchers in some of the shared tasks such as spontaneous speech recognition and understanding, voice data information extraction, and machine translation. The purpose of this special issue was to invite speech and language researchers to communicate with each other, and with the machine learning community on the latest machine learning advances in their work. The call for papers was met with great enthusiasm from the speech and natural language community. Thirty six submissions were received; each paper was reviewed by at least three reviewers. Only ten papers were selected re∞ecting not only some of the best work on machine learning in the areas of natural language and spoken language processing but also what we view as a collection of papers that represent current trends in these areas of research both from the perspective of

In the era of big data and machine learning, data privacy has become an increasingly critical concern. Traditional centralized data processing methods pose significant risks related to data leakage, unauthorized access, and regulatory non-compliance. Federated Learning (FL) and Distributed Databases (DD) have emerged as promising solutions to address these challenges. Federated Learning enables collaborative model training across decentralized devices or servers while keeping data localized, thus preserving privacy. Similarly, Distributed Databases facilitate secure data storage and access across multiple nodes, reducing the risks associated with single points of failure. This paper explores the synergy between Federated Learning and Distributed Databases in enhancing data privacy, providing a comprehensive review of the current literature, methodologies, and implementations. We examine key protocols, frameworks, and security mechanisms that support privacy-preserving machine learning. Through detailed analysis and case studies, we present the effectiveness of integrated FL-DD systems in real- world scenarios such as healthcare, finance, and IoT. We conclude with a discussion on challenges, future research directions, and the implications of combining FL and DD for building secure, scalable, and privacy-aware data systems. Keywords-Federated Learning, Distributed Databases, Data Privacy, Machine Learning, Secure Computation, Data Decentralization, Privacy- Preserving Models, IoT Security, Blockchain, Data Governance

Machine Learning (ML) and Deep Learning (DL) techniques have achieved remarkable results in many domains resulting in excellent performances in fields including Natural Language Processing (NLP), Computer Vision (CV) and Speech Recognition (SR). In light of these achievements, the application of ML algorithms to the industry world is appealing and very interesting results have already been produced by the scientific community. However, industrial data are often significantly different from those commonly employed in the disciplines where DL has had a larger impact. Industrial datasets are often limited in size and are constituted by large amounts of unlabelled data. In contrast, in CV and NLP, large deep neural networks can benefit from large labelled datasets, that can thus be used to train DL models in a fully-supervised fashion. In this work, we adopt a self-supervised approach to process heterogeneous dataset characterized by a relatively large number of unlabelled data and few labelled ones. We apply the proposed approach to two industrial time-series data and show that leveraging unlabelled data results in improved performance on classification downstream tasks.

While machine learning and artificial intelligence have long been applied in networking research, the bulk of such works has focused on supervised learning. Recently there has been a rising trend of employing unsupervised machine learning using unstructured raw network data to improve network performance and provide services such as traffic engineering, anomaly detection, Internet traffic classification, and quality of service optimization. The interest in applying unsupervised learning techniques in networking emerges from their great success in other fields such as computer vision, natural language processing, speech recognition, and optimal control (e.g., for developing autonomous self-driving cars). Unsupervised learning is interesting since it can unconstrain us from the need of labeled data and manual handcrafted feature engineering thereby facilitating flexible, general, and automated methods of machine learning. The focus of this survey paper is to provide an overview of the applications of unsupervised learning in the domain of networking. We provide a comprehensive survey highlighting the recent advancements in unsupervised learning techniques and describe their applications for various learning tasks in the context of networking. We also provide a discussion on future directions and open research issues, while also identifying potential pitfalls. While a few survey papers focusing on the applications of machine learning in networking have previously been published, a survey of similar scope and breadth is missing in literature. Through this paper, we advance the state of knowledge by carefully synthesizing the insights from these survey papers while also providing contemporary coverage of recent advances.

Federated Learning (FL) has transformed AI applications in healthcare by enabling collaborative model training across multiple institutions while preserving patient data privacy. Despite its advantages, FL remains susceptible to security vulnerabilities, including model inversion attacks, adversarial data poisoning, and communication inefficiencies, necessitating enhanced privacy-preserving mechanisms. In response, this study introduces Privacy-Preserving Federated Learning (PPFL), an advanced FL framework integrating Secure Multi-Party Computation (SMPC), Differential Privacy (DP), and Homomorphic Encryption (HE) to ensure data confidentiality while maintaining computational efficiency. I rigorously evaluate PPFL using Federated Averaging (FedAvg), Secure Aggregation (SecAgg), and Differentially Private Stochastic Gradient Descent (DP-SGD) across real-world healthcare datasets. The results indicate that this approach achieves up to an 85% reduction in model inversion attack success rates, enhances privacy efficiency by 30%, and maintains accuracy retention between 95.2% and 98.3%, significantly improving security without compromising model performance. Furthermore, comparative visual analyses illustrate trade-offs between privacy and accuracy, scalability trends, and computational overhead. This study also explores scalability challenges, computational trade-offs, and real-world deployment considerations in multi-institutional hospital networks, paving the way for secure, scalable, and privacy-preserving AI adoption in healthcare environments.