arXiv subjects

Hana Khamfroush

Publications and source records attributed to Hana Khamfroush.

14 recordsLinked to original sources

arXiv2026

One Round Is All You Need: Analytic Federated Learning for Task-Heterogeneous Multi-Label Medical Image Classification

Federated learning (FL) enables multiple clinical institutions to collaboratively train a shared disease classifier without centralizing patient data. In practice, however, each institution annotates only the pathologies within its area of expertise, so the federation operates under task heterogeneity: each client holds labels for a strict subset of the target disease categories while the remaining classes are entirely unobserved at that site. Existing gradient-based FL methods fail under this setting because they require hundreds of communication rounds to converge and because missing class labels introduce systematic false-negative bias that the model cannot correct without a principled mechanism. We propose an analytic federated learning framework for multi-label medical image classification under task heterogeneity. The proposed method replaces iterative gradient optimization with three closed-form operations: a balanced label projection that neutralizes class-imbalance bias by normalizing positive and negative contributions to equal total mass; a per-class absolute aggregation law that independently assembles the optimal ridge-regression classifier for each disease category from the sufficient statistics uploaded by its annotating clients; and an optional analytic pseudo-label refinement round that propagates missing-class knowledge from a confidence-filtered teacher classifier to non-annotating clients. The entire procedure requires at most two communication rounds, irrespective of the degree of task heterogeneity or the number of participating clients. Experiments on ChestXray14 under four progressively severe missing-class configurations demonstrate that the proposed method consistently outperforms the state-of-the-art federated multi-label method FedMLP by up to 18.44 BACC points and 13.24 AUC points, while reducing the communication.

cs.LG↗

arXiv2025

Semi-Supervised Federated Multi-Label Feature Selection with Fuzzy Information Measures

Multi-label feature selection (FS) reduces the dimensionality of multi-label data by removing irrelevant, noisy, and redundant features, thereby boosting the performance of multi-label learning models. However, existing methods typically require centralized data, which makes them unsuitable for distributed and federated environments where each device/client holds its own local dataset. Additionally, federated methods often assume that clients have labeled data, which is unrealistic in cases where clients lack the expertise or resources to label task-specific data. To address these challenges, we propose a Semi-Supervised Federated Multi-Label Feature Selection method, called SSFMLFS, where clients hold only unlabeled data, while the server has limited labeled data. SSFMLFS adapts fuzzy information theory to a federated setting, where clients compute fuzzy similarity matrices and transmit them to the server, which then calculates feature redundancy and feature-label relevancy degrees. A feature graph is constructed by modeling features as vertices, assigning relevancy and redundancy degrees as vertex weights and edge weights, respectively. PageRank is then applied to rank the features by importance. Extensive experiments on five real-world datasets from various domains, including biology, images, music, and text, demonstrate that SSFMLFS outperforms other federated and centralized supervised and semi-supervised approaches in terms of three different evaluation metrics in non-IID data distribution setting.

cs.LG↗

arXiv2025

Embedded Federated Feature Selection with Dynamic Sparse Training: Balancing Accuracy-Cost Tradeoffs

Federated Learning (FL) enables multiple resource-constrained edge devices with varying levels of heterogeneity to collaboratively train a global model. However, devices with limited capacity can create bottlenecks and slow down model convergence. One effective approach to addressing this issue is to use an efficient feature selection method, which reduces overall resource demands by minimizing communication and computation costs, thereby mitigating the impact of struggling nodes. Existing federated feature selection (FFS) methods are either considered as a separate step from FL or rely on a third party. These approaches increase computation and communication overhead, making them impractical for real-world high-dimensional datasets. To address this, we present \textit{Dynamic Sparse Federated Feature Selection} (DSFFS), the first innovative embedded FFS that is efficient in both communication and computation. In the proposed method, feature selection occurs simultaneously with model training. During training, input-layer neurons, their connections, and hidden-layer connections are dynamically pruned and regrown, eliminating uninformative features. This process enhances computational efficiency on devices, improves network communication efficiency, and boosts global model performance. Several experiments are conducted on nine real-world datasets of varying dimensionality from diverse domains, including biology, image, speech, and text. The results under a realistic non-iid data distribution setting show that our approach achieves a better trade-off between accuracy, computation, and communication costs by selecting more informative features compared to other state-of-the-art FFS methods.

cs.LG↗

arXiv2025

DILEMMA: Joint LLM Quantization and Distributed LLM Inference Over Edge Computing Systems

With a recent trend of using Large Language Models (LLMs) for different applications within smart cities, there is a need for pushing these models toward the edge of network while still preserving their performance. Edge Computing (EC) as a physically closer computing resource to the end users can help to reduce the communication delay for serving end users' tasks for LLM-dependent services. However, EC servers have limited capacity in terms of communication, computation, and storage capacity. This paper introduces DILEMMA, a novel framework addressing the challenges of deploying LLMs in EC systems by jointly optimizing layer placement and layer quantization in EC systems. DILEMMA formulates an Integer Linear Programming problem to minimize total inference delay while ensuring acceptable LLM performance levels, leveraging layer-wise quantization and knowledge distillation for LLM performance control. Experimental evaluations on OPT-350 model using the SQuAD dataset demonstrate that DILEMMA achieves a quantization ratio of up to 12.75% while preserving model loss, highlighting its effectiveness in resource-constrained environments.

cs.LG↗

arXiv2024

FMLFS: A Federated Multi-Label Feature Selection Based on Information Theory in IoT Environment

In certain emerging applications such as health monitoring wearable and traffic monitoring systems, Internet-of-Things (IoT) devices generate or collect a huge amount of multi-label datasets. Within these datasets, each instance is linked to a set of labels. The presence of noisy, redundant, or irrelevant features in these datasets, along with the curse of dimensionality, poses challenges for multi-label classifiers. Feature selection (FS) proves to be an effective strategy in enhancing classifier performance and addressing these challenges. Yet, there is currently no existing distributed multi-label FS method documented in the literature that is suitable for distributed multi-label datasets within IoT environments. This paper introduces FMLFS, the first federated multi-label feature selection method. Here, mutual information between features and labels serves as the relevancy metric, while the correlation distance between features, derived from mutual information and joint entropy, is utilized as the redundancy measure. Following aggregation of these metrics on the edge server and employing Pareto-based bi-objective and crowding distance strategies, the sorted features are subsequently sent back to the IoT devices. The proposed method is evaluated through two scenarios: 1) transmitting reduced-size datasets to the edge server for centralized classifier usage, and 2) employing federated learning with reduced-size datasets. Evaluation across three metrics - performance, time complexity, and communication cost - demonstrates that FMLFS outperforms five other comparable methods in the literature and provides a good trade-off on three real-world datasets.

cs.LG↗

arXiv2024

Enhancing IoT Security: A Novel Feature Engineering Approach for ML-Based Intrusion Detection Systems

The integration of Internet of Things (IoT) applications in our daily lives has led to a surge in data traffic, posing significant security challenges. IoT applications using cloud and edge computing are at higher risk of cyberattacks because of the expanded attack surface from distributed edge and cloud services, the vulnerability of IoT devices, and challenges in managing security across interconnected systems leading to oversights. This led to the rise of ML-based solutions for intrusion detection systems (IDSs), which have proven effective in enhancing network security and defending against diverse threats. However, ML-based IDS in IoT systems encounters challenges, particularly from noisy, redundant, and irrelevant features in varied IoT datasets, potentially impacting its performance. Therefore, reducing such features becomes crucial to enhance system performance and minimize computational costs. This paper focuses on improving the effectiveness of ML-based IDS at the edge level by introducing a novel method to find a balanced trade-off between cost and accuracy through the creation of informative features in a two-tier edge-user IoT environment. A hybrid Binary Quantum-inspired Artificial Bee Colony and Genetic Programming algorithm is utilized for this purpose. Three IoT intrusion detection datasets, namely NSL-KDD, UNSW-NB15, and BoT-IoT, are used for the evaluation of the proposed approach.

cs.CR↗

arXiv2023

Multimodal Multiple Federated Feature Construction Method for IoT Environments

The fast development of Internet-of-Things (IoT) devices and applications has led to vast data collection, potentially containing irrelevant, noisy, or redundant features that degrade learning model performance. These collected data can be processed on either end-user devices (clients) or edge/cloud server. Feature construction is a pre-processing technique that can generate discriminative features and reveal hidden relationships between original features within a dataset, leading to improved performance and reduced computational complexity of learning models. Moreover, the communication cost between clients and edge/cloud server can be minimized in situations where a dataset needs to be transmitted for further processing. In this paper, the first federated feature construction (FFC) method called multimodal multiple FFC (MMFFC) is proposed by using multimodal optimization and gravitational search programming algorithm. This is a collaborative method for constructing multiple high-level features without sharing clients' datasets to enhance the trade-off between accuracy of the trained model and overall communication cost of the system, while also reducing computational complexity of the learning model. We analyze and compare the accuracy-cost trade-off of two scenarios, namely, 1) MMFFC federated learning (FL), using vanilla FL with pre-processed datasets on clients and 2) MMFFC centralized learning, transferring pre-processed datasets to an edge server and using centralized learning model. The results on three datasets for the first scenario and eight datasets for the second one demonstrate that the proposed method can reduce the size of datasets for about 60$\%$, thereby reducing communication cost and improving accuracy of the learning models tested on almost all datasets.

cs.NI↗

arXiv2021

A Modified SEIR Model for the Spread of COVID-19 Considering Different Vaccine Types

The COVID-19 pandemic has influenced the lives of people globally. In the past year many researchers have proposed different models and approaches to explore in what ways the spread of the disease could be mitigated. One of the models that have been used a great deal is the Susceptible-Exposed-Infectious-Recovered (SEIR) model. Some researchers have modified the traditional SEIR model, and proposed new versions of it. However, to the best of our knowledge, the state-of-the-art papers have not considered the effect of different vaccine types, meaning single shot and double shot vaccines, in their SEIR model. In this paper, we propose a modified version of the SEIR model which takes into account the effect of different vaccine types. We compare how different policies for the administration of the vaccine can influence the rate at which people are exposed to the disease, get infected, recover, and pass away. Our results suggest that taking the double shot vaccine such as Pfizer-BioNTech and Moderna does a better job at mitigating the spread and fatality rate of the disease compared to the single shot vaccine, due to its higher efficacy.

q-bio.PE↗

arXiv2021

QoS-Aware Placement of Deep Learning Services on the Edge with Multiple Service Implementations

Mobile edge computing pushes computationally-intensive services closer to the user to provide reduced delay due to physical proximity. This has led many to consider deploying deep learning models on the edge -- commonly known as edge intelligence (EI). EI services can have many model implementations that provide different QoS. For instance, one model can perform inference faster than another (thus reducing latency) while achieving less accuracy when evaluated. In this paper, we study joint service placement and model scheduling of EI services with the goal to maximize Quality-of-Servcice (QoS) for end users where EI services have multiple implementations to serve user requests, each with varying costs and QoS benefits. We cast the problem as an integer linear program and prove that it is NP-hard. We then prove the objective is equivalent to maximizing a monotone increasing, submodular set function and thus can be solved greedily while maintaining a (1-1/e)-approximation guarantee. We then propose two greedy algorithms: one that theoretically guarantees this approximation and another that empirically matches its performance with greater efficiency. Finally, we thoroughly evaluate the proposed algorithm for making placement and scheduling decisions in both synthetic and real-world scenarios against the optimal solution and some baselines. In the real-world case, we consider real machine learning models using the ImageNet 2012 data-set for requests. Our numerical experiments empirically show that our more efficient greedy algorithm is able to approximate the optimal solution with a 0.904 approximation on average, while the next closest baseline achieves a 0.607 approximation on average.

cs.NI↗

arXiv2020

Optimal Accuracy-Time Trade-off for Deep Learning Services in Edge Computing Systems

With the increasing demand for computationally intensive services like deep learning tasks, emerging distributed computing platforms such as edge computing (EC) systems are becoming more popular. Edge computing systems have shown promising results in terms of latency reduction compared to the traditional cloud systems. However, their limited processing capacity imposes a trade-off between the potential latency reduction and the achieved accuracy in computationally-intensive services such as deep learning-based services. In this paper, we focus on finding the optimal accuracy-time trade-off for running deep learning services in a three-tier EC platform where several deep learning models with different accuracy levels are available. Specifically, we cast the problem as an Integer Linear Program, where optimal task scheduling decisions are made to maximize overall user satisfaction in terms of accuracy-time trade-off. We prove that our problem is NP-hard and then provide a polynomial constant-time greedy algorithm, called GUS, that is shown to attain near-optimal results. Finally, upon vetting our algorithmic solution through numerical experiments and comparison with a set of heuristics, we deploy it on a test-bed implemented to measure for real-world results. The results of both numerical analysis and real-world implementation show that GUS can outperform the baseline heuristics in terms of the average percentage of satisfied users by a factor of at least 50%.

cs.DC↗

arXiv2020

Smart Advertisement for Maximal Clicks in Online Social Networks Without User Data

Click-through rate (CTR) prediction of advertisements on online social network platforms to optimize advertising is of much interest. Prior works build machine learning models that take a user-centric approach in terms of training -- using predominantly user data to classify whether a user will click on an advertisement or not. While this approach has proven effective, it is inaccessible to most entities and relies heavily on user data. To accommodate for this, we first consider a large set of advertisement data on Facebook and use natural language processing (NLP) to extract key concepts that we call conceptual nodes. To predict the value of CTR for a combination of conceptual nodes, we use the advertisement data to train four machine learning (ML) models. We then cast the problem of finding the optimal combination of conceptual nodes as an optimization problem. Given a certain budget $k$, we are interested in finding the optimal combination of conceptual nodes that maximize the CTR. A discussion of the hardness and possible NP-hardness of the optimization problem is provided. Then, we propose a greedy algorithm and a genetic algorithm to find near-optimal combinations of conceptual nodes in polynomial time, with the genetic algorithm nearly matching the optimal solution. We observe that Decision Tree Regressor and Random Forest Regressor exhibit the highest Pearson correlation coefficients w.r.t. click predictions and real click values. Additionally, we find that the conceptual nodes of "politics", "celebrity", and "organization" are notably more influential than other considered conceptual nodes.

cs.SI↗

arXiv2020

Automatic Query Optimization for Retrieving Traffic Tweets

Twitter, like many social media and data brokering companies, makes their data available through a search API (application programming interface). In addition to filtering results by date and location, researchers can search for tweets with specific content with a boolean text query, using {\it AND}, {\it OR}, and {\it NOT} operators to select the combinations of phrases which must, or must not, appear in matching tweets. This boolean text search system is not at all unique to Twitter and is found in many different contexts, including academic, legal, and medical databases, however it is stretched to its limits in Twitter's use case because of the relative volume and brevity of tweets. In addition, the semi-automated use of such systems was well studied under the topic of Information Retrieval during the 1980s and 1990s, however the study of such systems has greatly declined since that time. As such, we propose updated methods for automatically selecting and refining complex boolean search queries that can isolate relevant results with greater specificity and completeness. Furthermore, we present preliminary results of using an optimized query to collect a sample of traffic-incident-related tweets, along with the results of manually classifying and analyzing them.

cs.SI↗

arXiv2019

Meeting QoS of Users in a Edge to Cloud Platform via Optimally Placing Services and Scheduling Tasks

This paper considers the problem of service placement and task scheduling on a three-tiered edge-to-cloud platform when user requests must be met by a certain deadline. Time-sensitive applications (e.g., augmented reality, gaming, real-time video analysis) have tight constraints that must be met. With multiple possible computation centers, the "where" and "when" of solving these requests becomes paramount when meeting their deadlines. We formulate the problem of meeting users' deadlines while minimizing the total cost of the edge-to-cloud service provider as an Integer Linear Programming (ILP) problem. We show the NP-hardness of this problem, and propose two heuristics based on making decisions on a local vs global scale. We vary the number of users, the QoS constraint, and the cost difference between remote cloud and cloudlets(edge clouds), and run multiple Monte-Carlo runs for each case. Our simulation results show that the proposed heuristics are performing close to optimal while reducing the complexity.

cs.NI↗

arXiv2019

On Fundamental Bounds of Failure Identifiability by Boolean Network Tomography

Boolean network tomography is a powerful tool to infer the state (working/failed) of individual nodes from path-level measurements obtained by egde-nodes. We consider the problem of optimizing the capability of identifying network failures through the design of monitoring schemes. Finding an optimal solution is NP-hard and a large body of work has been devoted to heuristic approaches providing lower bounds. Unlike previous works, we provide upper bounds on the maximum number of identifiable nodes, given the number of monitoring paths and different constraints on the network topology, the routing scheme, and the maximum path length. These upper bounds represent a fundamental limit on identifiability of failures via Boolean network tomography. Our analysis provides insights on how to design topologies and related monitoring schemes to achieve the maximum identifiability under various network settings. Through analysis and experiments we demonstrate the tightness of the bounds and efficacy of the design insights for engineered as well as real networks.

cs.NI↗