About

Stratis Ioannidis is a professor in the Department of Electrical and Computer Engineering at Northeastern University College of Engineering. His research focuses on machine learning and optimization, with particular emphasis on robust machine learning at the edge, distributed learning, and data-centric networks. He has been recognized with awards such as the 2025 Faculty Research Team Award – iSUPER Impact Engine, the 2025 Søren Buus Outstanding Research Award, and the 2021 College of Engineering Faculty Fellow Fellowship. He has led and contributed to multiple research projects funded by the National Science Foundation, including initiatives on machine learning at the edge, distributed optimization, and data-centric networks. His work has earned him a spot among the top 2% of most-cited scientists worldwide, as recognized by Stanford University. Ioannidis has also been involved in developing innovative solutions for content distribution and network caching, holding patents in these areas. His contributions extend to advancing the understanding and application of machine learning, communication, control, signal processing, and algorithms in engineering.

Research topics

• Computer Science
• Artificial Intelligence
• Computer Security
• Medicine
• Pathology
• Radiology
• Telecommunications
• Computer network

Selected publications

• SSplain: Sparse and Smooth Explainer for Retinopathy of Prematurity Classification

  2026-03-06

  article

  Neural networks are frequently used in medical diagnosis. However, due to their black-box nature, model explainers are used to help clinicians understand better and trust model outputs. This paper introduces an explainer method for classifying Retinopathy of Prematurity (ROP) from fundus images. Previous methods fail to generate explanations that preserve input image structures such as smoothness and sparsity. We introduce Sparse and Smooth Explainer (SSplain), a method that generates pixel-wise explanations while preserving image structures by enforcing smoothness and sparsity. This results in realistic explanations to enhance the understanding of the given black-box model. To achieve this goal, we define an optimization problem with combinatorial constraints and solve it using the Alternating Direction Method of Multipliers (ADMM). Experimental results show that SSplain outperforms commonly used explainers in terms of both post-hoc accuracy and smoothness analyses. Additionally, SSplain identifies features that are consistent with domain-understandable features that clinicians consider as discriminative factors for ROP. We also show SSplain’s generalization by applying it to additional publicly available datasets. Code is available at https://github.com/neu-spiral/SSplain.

  PublisherDOI

• Fair Concurrent Training of Multiple Models in Federated Learning

  IEEE Transactions on Networking · 2025-06-23 · 5 citations

  article

  Federated learning (FL) enables collaborative learning across multiple clients. In most FL work, all clients train a single learning task. However, the recent proliferation of FL applications may increasingly require multiple FL tasks to be trained simultaneously, sharing clients’ computing resources, which we call Multiple-Model Federated Learning (MMFL). Current MMFL algorithms use naïve average-based client-task allocation schemes that often lead to unfair performance when FL tasks have heterogeneous difficulty levels, as the more difficult tasks may need more client participation to train effectively. Furthermore, in the MMFL setting, we face a further challenge that some clients may prefer training specific tasks to others, and may not even be willing to train other tasks, e.g., due to high computational costs, which may exacerbate unfairness in training outcomes across tasks. We address both challenges by firstly designing FedFairMMFL, a difficulty-aware algorithm that dynamically allocates clients to tasks in each training round, based on the tasks’ current performance levels. We provide guarantees on the resulting task fairness and FedFairMMFL’s convergence rate. We then propose novel auction designs that incentivizes clients to train multiple tasks, so as to fairly distribute clients’ training efforts across the tasks, and extend our convergence guarantees to this setting. We finally evaluate our algorithm with multiple sets of learning tasks on real world datasets, showing that our algorithm improves fairness by improving the final model accuracy and convergence speed of the worst performing tasks, while maintaining the average accuracy across tasks.

  PublisherDOI

• Mind the Gap: Delayed Label Bias-Variance Tradeoffs in Predicting Likelihood of Nonpayment

  2025-08-03

  articleOpen accessSenior author

  The purpose of an online electronic-payment risk detection system is to prevent leakage, i.e., the loss of revenue that occurs when users fail to pay for services or when transactions are reversed.Nonpayment prediction models are trained on datasets comprising of features available when the model is triggered and the corresponding nonpayment labels.The latter are typically only observed several weeks or even months later.Furthermore, behavior indicative of future nonpayment is highly non-stationary, and the true model may drift significantly in the gap between trigger events and label collection.To address these challenges, we use post-transaction signals to generate pseudo-labels, i.e., short-term proxies [23] or surrogate-indices [33].Our framework attains a favorable tradeoff between ameliorating bias due to drift and introducing variance due to pseudo-label noise, as demonstrated by both offline and online experiments on several nonpayment-detection systems at Meta.Our deployment on live user traffic yields a statistically significant improvement in revenue, accounting also for leakage.

  PublisherOA PDFDOI

• Spectral Survival Analysis

  2025-08-03 · 1 citations

  preprintOpen accessSenior author

  Survival analysis is widely deployed in a diverse set of fields, including healthcare, business, ecology, etc. The Cox Proportional Hazard (CoxPH) model is a semi-parametric model often encountered in the literature. Despite its popularity, wide deployment, and numerous variants, scaling CoxPH to large datasets and deep architectures poses a challenge, especially in the high-dimensional regime. We identify a fundamental connection between rank regression and the CoxPH model: this allows us to adapt and extend the so-called spectral method for rank regression to survival analysis. Our approach is versatile, naturally generalizing to several CoxPH variants, including deep models. We empirically verify our method's scalability on multiple real-world high-dimensional datasets; our method outperforms legacy methods w.r.t. predictive performance and efficiency.

  PublisherOA PDFDOI

• Empowering Federated Learning With Implicit Gossiping: Mitigating Connection Unreliability Amidst Unknown and Arbitrary Dynamics

  IEEE Transactions on Signal Processing · 2025-01-01 · 1 citations

  article

  Federated learning is a popular distributed learning approach for training a machine learning model without disclosing raw data. It consists of a parameter server and a possibly large collection of clients (e.g., in cross-device federated learning) that may operate in congested and changing environments. In this paper, we study federated learning in the presence of stochastic and dynamic communication failures wherein the uplink between the parameter server and client <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$i$</tex-math></inline-formula> is on with <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">unknown</i> probability <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$p_{i}^{t}$</tex-math></inline-formula> in round <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$t$</tex-math></inline-formula>. Furthermore, we allow the dynamics of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$p_{i}^{t}$</tex-math></inline-formula> to be <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">arbitrary</i>. We first demonstrate that when the <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$p_{i}^{t}$</tex-math></inline-formula>'s vary across clients, the most widely adopted federated learning algorithm, Federated Average (FedAvg), experiences significant bias. To address this observation, we propose Federated Postponed Broadcast (FedPBC), a simple variant of FedAvg. It differs from FedAvg in that the parameter server postpones broadcasting the global model to the clients with active uplinks till the end of each training round. Despite uplink failures, we show that FedPBC converges to a stationary point of the original non-convex objective. On the technical front, postponing the global model broadcasts enables implicit gossiping among the clients with active links in round <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$t$</tex-math></inline-formula>. In spite of the time-varying nature of <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$p_{i}^{t}$</tex-math></inline-formula>, we can bound the perturbation of the global model dynamics using techniques to control gossip-type information mixing errors. Extensive experiments have been conducted on real-world datasets over diversified unreliable uplink patterns to corroborate our analysis.

  PublisherDOI

• Learning Set Functions with Implicit Differentiation

  Proceedings of the AAAI Conference on Artificial Intelligence · 2025-04-11

  articleOpen accessSenior author

  A recent work introduces the problem of learning set functions from data generated by a so-called optimal subset oracle. Their approach approximates the underlying utility function with an energy-based model, whose parameters are estimated via mean-field variational inference. This approximation reduces to fixed point iterations; however, as the number of iterations increases, automatic differentiation quickly becomes computationally prohibitive due to the size of the Jacobians that are stacked during backpropagation. We address this challenge with implicit differentiation and examine the convergence conditions for the fixed-point iterations. We empirically demonstrate the efficiency of our method on synthetic and real-world subset selection applications including product recommendation, set anomaly detection and compound selection tasks.

  PublisherOA PDFDOI

• Online Two-Stage Submodular Maximization

  ArXiv.org · 2025-10-22

  preprintOpen access

  Given a collection of monotone submodular functions, the goal of Two-Stage Submodular Maximization (2SSM) [Balkanski et al., 2016] is to restrict the ground set so an objective selected u.a.r. from the collection attains a high maximal value, on average, when optimized over the restricted ground set. We introduce the Online Two-Stage Submodular Maximization (O2SSM) problem, in which the submodular objectives are revealed in an online fashion. We study this problem for weighted threshold potential functions, a large and important subclass of monotone submodular functions that includes influence maximization, data summarization, and facility location, to name a few. We design an algorithm that achieves sublinear $(1 - 1/e)^2$-regret under general matroid constraints and $(1 - 1/e)(1-e^{-k}k^k/k!)$-regret in the case of uniform matroids of rank $k$; the latter also yields a state-of-the-art bound for the (offline) 2SSM problem. We empirically validate the performance of our online algorithm with experiments on real datasets.

  PublisherOA PDFDOI

• H-SPLID: HSIC-based Saliency Preserving Latent Information Decomposition

  ArXiv.org · 2025-10-23

  preprintOpen accessSenior author

  We introduce H-SPLID, a novel algorithm for learning salient feature representations through the explicit decomposition of salient and non-salient features into separate spaces. We show that H-SPLID promotes learning low-dimensional, task-relevant features. We prove that the expected prediction deviation under input perturbations is upper-bounded by the dimension of the salient subspace and the Hilbert-Schmidt Independence Criterion (HSIC) between inputs and representations. This establishes a link between robustness and latent representation compression in terms of the dimensionality and information preserved. Empirical evaluations on image classification tasks show that models trained with H-SPLID primarily rely on salient input components, as indicated by reduced sensitivity to perturbations affecting non-salient features, such as image backgrounds. Our code is available at https://github.com/neu-spiral/H-SPLID.

  PublisherOA PDFDOI

• Enabling Flexible Multi-LLM Integration for Scalable Knowledge Aggregation

  ArXiv.org · 2025-05-28

  preprintOpen access

  Large language models (LLMs) have shown remarkable promise but remain challenging to continually improve through traditional finetuning, particularly when integrating capabilities from other specialized LLMs. Popular methods like ensemble and weight merging require substantial memory and struggle to adapt to changing data environments. Recent efforts have transferred knowledge from multiple LLMs into a single target model; however, they suffer from interference and degraded performance among tasks, largely due to limited flexibility in candidate selection and training pipelines. To address these issues, we propose a framework that adaptively selects and aggregates knowledge from diverse LLMs to build a single, stronger model, avoiding the high memory overhead of ensemble and inflexible weight merging. Specifically, we design an adaptive selection network that identifies the most relevant source LLMs based on their scores, thereby reducing knowledge interference. We further propose a dynamic weighted fusion strategy that accounts for the inherent strengths of candidate LLMs, along with a feedback-driven loss function that prevents the selector from converging on a single subset of sources. Experimental results demonstrate that our method can enable a more stable and scalable knowledge aggregation process while reducing knowledge interference by up to 50% compared to existing approaches. Code is avaliable at https://github.com/ZLKong/LLM_Integration

  PublisherOA PDFDOI

• DITTO: DIgital Twins for Testing and Optimizing Wireless Decisions

  2024-08-12 · 3 citations

  articleSenior author

  Digital Twins (DTs) are powerful tools for decision making that mirror real-world systems and continuous interactions between them. We study on the application and capabilities of DTs in the realm of wireless communications, using two leading wireless communication tools: Wireless InSite and Sionna. Specifically, we compare the fidelity of the two wireless communication tools by measuring several metrics with a real-life dataset. Our comprehensive analysis aims to determine the capabilities of Digital Twins in the context of wireless communications, offering valuable insights for future researchers in the field.

  PublisherDOI

Recent grants

• SaTC: CORE: Small: Massively Scalable Secure Computation Infrastructure Using FPGAs

  NSF · $500k · 2017–2022

• Collaborative Research: CNS Core: Medium: Data-Centric Networks for Distributed Learning

  NSF · $550k · 2021–2025

• CAREER: Leveraging Sparsity in Massively Distributed Optimization

  NSF · $459k · 2018–2025

• RTML: Large: Efficient and Adaptive Real-Time Learning for Next Generation Wireless Systems

  NSF · $1.2M · 2019–2023

• NeTS: Small: Caching Networks with Optimality Guarantees

  NSF · $500k · 2017–2022

Frequent coauthors

• Jayashree Kalpathy‐Cramer

  University of Colorado Anschutz Medical Campus

  93 shared

• Jennifer Dy

  Northeastern University

  72 shared

• Michael F. Chiang

  National Eye Institute

  69 shared

• Deniz Erdoğmuş

  Northeastern University

  57 shared

• J. Peter Campbell

  54 shared

• Susan Ostmo

  Oregon Health & Science University

  54 shared

• Kaushik Chowdhury

  Boston University

  45 shared

• James M. Brown

  38 shared

Labs

• Northeastern University College of Engineering - Stratis Ioannidis LabPI

Awards & honors

• 2025 Faculty Research Team Award – iSUPER Impact Engine
• 2025 Søren Buus Outstanding Research Award
• Fellowship, Greek Diaspora Fellowship Program
• National Science Foundation CAREER Award