arXiv subjects

Pavlos S. Efraimidis

Publications and source records attributed to Pavlos S. Efraimidis.

At least 19 recordsLinked to original sources

arXiv2026

Robust and Explainable Bicuspid Aortic Valve Diagnosis Using Stacked Ensembles on Echocardiography

Transthoracic echocardiography (TTE) is the first-line imaging modality for diagnosing bicuspid aortic valve (BAV), yet diagnostic performance varies with operator expertise and image quality. We developed an explainable AI model that distinguishes BAV from tricuspid aortic valves (TAV) using routinely acquired parasternal long-axis (PLAX) cine loops. A multi-backbone video ensemble was trained and evaluated using a leakage-aware, stratified outer cross-validation protocol on $N{=}90$ patient studies (48 BAV, 42 TAV). Across fixed outer splits and 10 random seeds, the calibrated stacked ensemble achieved an outer-CV F1-score of $0.907$ and recall of $0.877$. Frame-level Grad-CAM localized salient evidence to the aortic root and leaflet plane, while globally aggregated SHAP values quantified each video backbone's contribution to the stacked prediction, enabling transparent, case-level auditability. These findings indicate that PLAX-based video ensembles can support reliable BAV/TAV classification from routine echocardiographic cine loops and may facilitate earlier detection in non-specialist or resource-limited clinical settings.

cs.LG↗

arXiv2026

Evaluating the Defense Potential of Machine Unlearning against Membership Inference Attacks

Membership Inference Attacks (MIAs) pose a significant privacy risk by enabling adversaries to determine if a specific data point was part of a model's training set. This work empirically investigates whether MU algorithms can function as a targeted, active defense mechanism, in scenarios where a privacy audit identifies specific classes or individuals as highly susceptible to MIAs post-training. By 'dulling' the model's categorical memory of these samples, the process effectively mitigates the membership signal and reduces the MIA success rate for the most vulnerable users. We evaluate the defense potential of three MU algorithms, Negative Gradient (neg grad), SCalable Remembering and Unlearning unBound (SCRUB), and Selective Fine-tuning and Targeted Confusion (SFTC), across four diverse datasets and three complexity-based model groups. Our findings reveal that MU can function as a countermeasure against MIAs, though its success is critically contingent on algorithm choice, model capacity, and a profound sensitivity to learning rates. While Negative Gradient often induces a generalized degradation of membership signals across both forget and retain set, SFTC identifies a critical ``divergence effect'' where targeted forgetting reinforces the membership signal of retained data. Conversely, SCRUB provides a more balanced defense with minimal collateral impact on MIA perspective.

cs.CR↗

arXiv2025

Large Language Models as Universal Predictors? An Empirical Study on Small Tabular Datasets

Large Language Models (LLMs), originally developed for natural language processing (NLP), have demonstrated the potential to generalize across modalities and domains. With their in-context learning (ICL) capabilities, LLMs can perform predictive tasks over structured inputs without explicit fine-tuning on downstream tasks. In this work, we investigate the empirical function approximation capability of LLMs on small-scale structured datasets for classification, regression and clustering tasks. We evaluate the performance of state-of-the-art LLMs (GPT-5, GPT-4o, GPT-o3, Gemini-2.5-Flash, DeepSeek-R1) under few-shot prompting and compare them against established machine learning (ML) baselines, including linear models, ensemble methods and tabular foundation models (TFMs). Our results show that LLMs achieve strong performance in classification tasks under limited data availability, establishing practical zero-training baselines. In contrast, the performance in regression with continuous-valued outputs is poor compared to ML models, likely because regression demands outputs in a large (often infinite) space, and clustering results are similarly limited, which we attribute to the absence of genuine ICL in this setting. Nonetheless, this approach enables rapid, low-overhead data exploration and offers a viable alternative to traditional ML pipelines in business intelligence and exploratory analytics contexts. We further analyze the influence of context size and prompt structure on approximation quality, identifying trade-offs that affect predictive performance. Our findings suggest that LLMs can serve as general-purpose predictive engines for structured data, with clear strengths in classification and significant limitations in regression and clustering.

cs.AI↗

arXiv2024

Federated Learning in Mobile Networks: A Comprehensive Case Study on Traffic Forecasting

The increasing demand for efficient resource allocation in mobile networks has catalyzed the exploration of innovative solutions that could enhance the task of real-time cellular traffic prediction. Under these circumstances, federated learning (FL) stands out as a distributed and privacy-preserving solution to foster collaboration among different sites, thus enabling responsive near-the-edge solutions. In this paper, we comprehensively study the potential benefits of FL in telecommunications through a case study on federated traffic forecasting using real-world data from base stations (BSs) in Barcelona (Spain). Our study encompasses relevant aspects within the federated experience, including model aggregation techniques, outlier management, the impact of individual clients, personalized learning, and the integration of exogenous sources of data. The performed evaluation is based on both prediction accuracy and sustainability, thus showcasing the environmental impact of employed FL algorithms in various settings. The findings from our study highlight FL as a promising and robust solution for mobile traffic prediction, emphasizing its twin merits as a privacy-conscious and environmentally sustainable approach, while also demonstrating its capability to overcome data heterogeneity and ensure high-quality predictions, marking a significant stride towards its integration in mobile traffic management systems.

cs.LG↗

arXiv2024

Federated Anomaly Detection for Early-Stage Diagnosis of Autism Spectrum Disorders using Serious Game Data

Early identification of Autism Spectrum Disorder (ASD) is considered critical for effective intervention to mitigate emotional, financial and societal burdens. Although ASD belongs to a group of neurodevelopmental disabilities that are not curable, researchers agree that targeted interventions during childhood can drastically improve the overall well-being of individuals. However, conventional ASD detection methods such as screening tests, are often costly and time-consuming. This study presents a novel semi-supervised approach for ASD detection using AutoEncoder-based Machine Learning (ML) methods due to the challenge of obtaining ground truth labels for the associated task. Our approach utilizes data collected manually through a serious game specifically designed for this purpose. Since the sensitive data collected by the gamified application are susceptible to privacy leakage, we developed a Federated Learning (FL) framework that can enhance user privacy without compromising the overall performance of the ML models. The framework is further enhanced with Fully Homomorphic Encryption (FHE) during model aggregation to minimize the possibility of inference attacks and client selection mechanisms as well as state-of-the-art aggregators to improve the model's predictive accuracy. Our results demonstrate that semi-supervised FL can effectively predict an ASD risk indicator for each case while simultaneously addressing privacy concerns.

cs.CY↗

arXiv2023

Intelligent Client Selection for Federated Learning using Cellular Automata

Federated Learning (FL) has emerged as a promising solution for privacy-enhancement and latency minimization in various real-world applications, such as transportation, communications, and healthcare. FL endeavors to bring Machine Learning (ML) down to the edge by harnessing data from million of devices and IoT sensors, thus enabling rapid responses to dynamic environments and yielding highly personalized results. However, the increased amount of sensors across diverse applications poses challenges in terms of communication and resource allocation, hindering the participation of all devices in the federated process and prompting the need for effective FL client selection. To address this issue, we propose Cellular Automaton-based Client Selection (CA-CS), a novel client selection algorithm, which leverages Cellular Automata (CA) as models to effectively capture spatio-temporal changes in a fast-evolving environment. CA-CS considers the computational resources and communication capacity of each participating client, while also accounting for inter-client interactions between neighbors during the client selection process, enabling intelligent client selection for online FL processes on data streams that closely resemble real-world scenarios. In this paper, we present a thorough evaluation of the proposed CA-CS algorithm using MNIST and CIFAR-10 datasets, while making a direct comparison against a uniformly random client selection scheme. Our results demonstrate that CA-CS achieves comparable accuracy to the random selection approach, while effectively avoiding high-latency clients.

cs.LG↗

arXiv2023

Towards Energy-Aware Federated Traffic Prediction for Cellular Networks

Cellular traffic prediction is a crucial activity for optimizing networks in fifth-generation (5G) networks and beyond, as accurate forecasting is essential for intelligent network design, resource allocation and anomaly mitigation. Although machine learning (ML) is a promising approach to effectively predict network traffic, the centralization of massive data in a single data center raises issues regarding confidentiality, privacy and data transfer demands. To address these challenges, federated learning (FL) emerges as an appealing ML training framework which offers high accurate predictions through parallel distributed computations. However, the environmental impact of these methods is often overlooked, which calls into question their sustainability. In this paper, we address the trade-off between accuracy and energy consumption in FL by proposing a novel sustainability indicator that allows assessing the feasibility of ML models. Then, we comprehensively evaluate state-of-the-art deep learning (DL) architectures in a federated scenario using real-world measurements from base station (BS) sites in the area of Barcelona, Spain. Our findings indicate that larger ML models achieve marginally improved performance but have a significant environmental impact in terms of carbon footprint, which make them impractical for real-world applications.

cs.LG↗

arXiv2023

Federated Learning for Early Dropout Prediction on Healthy Ageing Applications

The provision of social care applications is crucial for elderly people to improve their quality of life and enables operators to provide early interventions. Accurate predictions of user dropouts in healthy ageing applications are essential since they are directly related to individual health statuses. Machine Learning (ML) algorithms have enabled highly accurate predictions, outperforming traditional statistical methods that struggle to cope with individual patterns. However, ML requires a substantial amount of data for training, which is challenging due to the presence of personal identifiable information (PII) and the fragmentation posed by regulations. In this paper, we present a federated machine learning (FML) approach that minimizes privacy concerns and enables distributed training, without transferring individual data. We employ collaborative training by considering individuals and organizations under FML, which models both cross-device and cross-silo learning scenarios. Our approach is evaluated on a real-world dataset with non-independent and identically distributed (non-iid) data among clients, class imbalance and label ambiguity. Our results show that data selection and class imbalance handling techniques significantly improve the predictive accuracy of models trained under FML, demonstrating comparable or superior predictive performance than traditional ML models.

cs.LG↗

arXiv2023

Federated Learning for 5G Base Station Traffic Forecasting

Cellular traffic prediction is of great importance on the path of enabling 5G mobile networks to perform intelligent and efficient infrastructure planning and management. However, available data are limited to base station logging information. Hence, training methods for generating high-quality predictions that can generalize to new observations across diverse parties are in demand. Traditional approaches require collecting measurements from multiple base stations, transmitting them to a central entity and conducting machine learning operations using the acquire data. The dissemination of local observations raises concerns regarding confidentiality and performance, which impede the applicability of machine learning techniques. Although various distributed learning methods have been proposed to address this issue, their application to traffic prediction remains highly unexplored. In this work, we investigate the efficacy of federated learning applied to raw base station LTE data for time-series forecasting. We evaluate one-step predictions using five different neural network architectures trained with a federated setting on non-identically distributed data. Our results show that the learning architectures adapted to the federated setting yield equivalent prediction error to the centralized setting. In addition, preprocessing techniques on base stations enhance forecasting accuracy, while advanced federated aggregators do not surpass simpler approaches. Simulations considering the environmental impact suggest that federated learning holds the potential for reducing carbon emissions and energy consumption. Finally, we consider a large-scale scenario with synthetic data and demonstrate that federated learning reduces the computational and communication costs compared to centralized settings.

cs.LG↗

arXiv2023

Predicting Early Dropouts of an Active and Healthy Ageing App

In this work, we present a machine learning approach for predicting early dropouts of an active and healthy ageing app. The presented algorithms have been submitted to the IFMBE Scientific Challenge 2022, part of IUPESM WC 2022. We have processed the given database and generated seven datasets. We used pre-processing techniques to construct classification models that predict the adherence of users using dynamic and static features. We submitted 11 official runs and our results show that machine learning algorithms can provide high-quality adherence predictions. Based on the results, the dynamic features positively influence a model's classification performance. Due to the imbalanced nature of the dataset, we employed oversampling methods such as SMOTE and ADASYN to improve the classification performance. The oversampling approaches led to a remarkable improvement of 10\%. Our methods won first place in the IFMBE Scientific Challenge 2022.

cs.LG↗

arXiv2022

An Exact, Linear Time Barabási-Albert Algorithm

This paper presents the development of a new class of algorithms that accurately implement the preferential attachment mechanism of the Barabási-Albert (BA) model to generate scale-free graphs. Contrary to existing approximate preferential attachment schemes, our methods are exact in terms of the proportionality of the vertex selection probabilities to their degree and run in linear time with respect to the order of the generated graph. Our algorithms utilize a series of precise, diverse, weighted and unweighted random sampling steps to engineer the desired properties of the graph generator. We analytically show that they obey the definition of the original BA model that generates scale-free graphs and discuss their higher-order properties. The proposed methods additionally include options to manipulate one dimension of control over the joint inclusion of groups of vertices.

cs.DS↗

arXiv2022

Federated Neural Collaborative Filtering

In this work, we present a federated version of the state-of-the-art Neural Collaborative Filtering (NCF) approach for item recommendations. The system, named FedNCF, enables learning without requiring users to disclose or transmit their raw data. Data localization preserves data privacy and complies with regulations such as the GDPR. Although federated learning enables model training without local data dissemination, the transmission of raw clients' updates raises additional privacy issues. To address this challenge, we incorporate a privacy-preserving aggregation method that satisfies the security requirements against an honest but curious entity. We argue theoretically and experimentally that existing aggregation algorithms are inconsistent with latent factor model updates. We propose an enhancement by decomposing the aggregation step into matrix factorization and neural network-based averaging. Experimental validation shows that FedNCF achieves comparable recommendation quality to the original NCF system, while our proposed aggregation leads to faster convergence compared to existing methods. We investigate the effectiveness of the federated recommender system and evaluate the privacy-preserving mechanism in terms of computational cost.

cs.IR↗

arXiv2021

FedPOIRec: Privacy Preserving Federated POI Recommendation with Social Influence

With the growing number of Location-Based Social Networks, privacy preserving location prediction has become a primary task for helping users discover new points-of-interest (POIs). Traditional systems consider a centralized approach that requires the transmission and collection of users' private data. In this work, we present FedPOIRec, a privacy preserving federated learning approach enhanced with features from users' social circles for top-$N$ POI recommendations. First, the FedPOIRec framework is built on the principle that local data never leave the owner's device, while the local updates are blindly aggregated by a parameter server. Second, the local recommenders get personalized by allowing users to exchange their learned parameters, enabling knowledge transfer among friends. To this end, we propose a privacy preserving protocol for integrating the preferences of a user's friends after the federated computation, by exploiting the properties of the CKKS fully homomorphic encryption scheme. To evaluate FedPOIRec, we apply our approach into five real-world datasets using two recommendation models. Extensive experiments demonstrate that FedPOIRec achieves comparable recommendation quality to centralized approaches, while the social integration protocol incurs low computation and communication overhead on the user side.

cs.LG↗

arXiv2021

About Weighted Random Sampling in Preferential Attachment Models

The Barabási-Albert model is a popular scheme for creating scale-free graphs but has been previously shown to have ambiguities in its definition. In this paper we discuss a new ambiguity in the definition of the BA model by identifying the tight relation between the preferential attachment process and unequal probability random sampling. While the probability that each individual vertex is selected is set to be proportional to their degree, the model does not specify the joint probabilities that any tuple of $m$ vertices is selected together for $m>1$. We demonstrate the consequences using analytical, experimental, and empirical analyses and propose a concise definition of the model that addresses this ambiguity. Using the connection with unequal probability random sampling, we also highlight a confusion about the process via which nodes are selected on each time step, for which - despite being implicitly indicated in the original paper - current literature appears fragmented.

cs.DS↗

arXiv2021

Lexicographic Enumeration of Set Partitions

In this report, we summarize the set partition enumeration problems and thoroughly explain the algorithms used to solve them. These algorithms iterate through the partitions in lexicographic order and are easy to understand and implement in modern high-level programming languages, without recursive structures and jump logic. We show that they require linear space in respect to the set cardinality and advance the enumeration in constant amortized time. The methods discussed in this document are not novel. Our goal is to demonstrate the process of enumerating set partitions and highlight the ideas behind it. This work is an aid for learners approaching this enumeration problem and programmers undertaking the task of implementing it.

cs.DM↗

arXiv2015

Weighted Random Sampling over Data Streams

In this work, we present a comprehensive treatment of weighted random sampling (WRS) over data streams. More precisely, we examine two natural interpretations of the item weights, describe an existing algorithm for each case ([2, 4]), discuss sampling with and without replacement and show adaptations of the algorithms for several WRS problems and evolving data streams.

cs.DS↗

arXiv2012

On Money as a Means of Coordination between Network Packets

In this work, we apply a common economic tool, namely money, to coordinate network packets. In particular, we present a network economy, called PacketEconomy, where each flow is modeled as a population of rational network packets, and these packets can self-regulate their access to network resources by mutually trading their positions in router queues. Every packet of the economy has its price, and this price determines if and when the packet will agree to buy or sell a better position. We consider a corresponding Markov model of trade and show that there are Nash equilibria (NE) where queue positions and money are exchanged directly between the network packets. This simple approach, interestingly, delivers improvements even when fiat money is used. We present theoretical arguments and experimental results to support our claims.

cs.DS↗

arXiv2011

An exact and O(1) time heaviest and lightest hitters algorithm for sliding-window data streams

In this work we focus on the problem of finding the heaviest-k and lightest-k hitters in a sliding window data stream. The most recent research endeavours have yielded an epsilon-approximate algorithm with update operations in constant time with high probability and O(1/epsilon) query time for the heaviest hitters case. We propose a novel algorithm which for the first time, to our knowledge, provides exact, not approximate, results while at the same time achieves O(1) time with high probability complexity on both update and query operations. Furthermore, our algorithm is able to provide both the heaviest-k and the lightest-k hitters at the same time without any overhead. In this work, we describe the algorithm and the accompanying data structure that supports it and perform quantitative experiments with synthetic data to verify our theoretical predictions.

cs.DS↗