Animation and exchanges in the Petri Nets community

[Marco Paolieri <address@concealed>]

============================================
ACM TOMPECS (Transactions on Modeling and
Performance Evaluation of Computing Systems)
============================================

Special Issue:
Performance Evaluation of Federated Learning Systems
https://dl.acm.org/journal/tompecs/calls-for-papers


MOTIVATION
==========

Federated learning has recently emerged as a trendy privacy-preserving
approach for training machine learning models on data that is
scattered across multiple heterogeneous devices/clients. In federated
learning, clients iteratively compute updates to the machine learning
models on their local datasets. These updates are periodically
aggregated across clients, typically but not always with the help of a
central parameter server.

In many real-world applications of federated learning such as
connected-and-autonomous vehicles (CAVs), the underlying
distributed/decentralized systems on which federated learning
algorithms are executing suffer a wide degree of heterogeneity
including but not limited to data distributions, computation speeds,
and external local environments. Moreover, the clients in federated
learning systems are often resource-constrained edge or end devices
and may compete for common resources such as communication bandwidth.

Many federated learning algorithms have been proposed and analyzed
experimentally and theoretically, yet these only cover a limited range
of heterogeneity. In addition, running federated learning in
resource-constrained settings often presents complex and not
well-understood tradeoffs among various performance metrics, including
final accuracy, convergence rate, and resource consumption.


TOPICS
======

This special issue will focus on the performance evaluation of
federated learning systems. We solicit papers that include theoretical
models or numerical analysis of federated learning performance, as
well as system-oriented papers that evaluate implementations of
federated learning systems. Specific topics of interest include, but
are not limited to:

- Novel techniques for analyzing the convergence of federated
  learning algorithms
- Performance analysis of emerging federated learning paradigms, e.g.,
  personalized models, asynchronous learning, cache-enhanced learning
- Analysis of performance tradeoffs in federated learning systems
- Active client selection in federated learning
- Fairness metrics for federated learning systems
- Novel federated learning algorithms that aim to address system
  heterogeneity or other practical implementation challenges, e.g.,
  dynamic client availability
- Benchmark platforms that enable evaluation of multiple federated
  learning algorithms
- New federated learning algorithms or analysis frameworks motivated
  by specific applications, e.g., large language models or
  recommendation systems
- Experimental results from large-scale federated learning deployments


IMPORTANT DATES
===============

- Submissions deadline: April 22, 2024
- First-round review decisions: June 30, 2024
- Deadline for revision submissions: August 31, 2024
- Notification of final decisions: October 15, 2024
- Tentative publication: December 1, 2024


SUBMISSION INFORMATION
======================

Submissions should follow the standard ACM TOMPECS formatting
requirements:

https://dl.acm.org/journal/tompecs/author-guidelines#submission

We will use Manuscript Central (https://mc.manuscriptcentral.com/tompecs)
to handle submissions.


GUEST EDITORS
=============

- Carlee Joe-Wong, Carnegie Mellon University
  address@concealed

- Lili Su, Northeastern University
  address@concealed

-- 
To unsubscribe from this group and stop receiving emails from it, send an email to address@concealed.