Benchmarking Catastrophic Forgetting Mitigation Methods in Federated Time Series Forecasting

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Hallak, Khaled, Kem, Oudom
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915811713613824
author Hallak, Khaled
Kem, Oudom
author_facet Hallak, Khaled
Kem, Oudom
contents Catastrophic forgetting (CF) poses a persistent challenge in continual learning (CL), especially within federated learning (FL) environments characterized by non-i.i.d. time series data. While existing research has largely focused on classification tasks in vision domains, the regression-based forecasting setting prevalent in IoT and edge applications remains underexplored. In this paper, we present the first benchmarking framework tailored to investigate CF in federated continual time series forecasting. Using the Beijing Multi-site Air Quality dataset across 12 decentralized clients, we systematically evaluate several CF mitigation strategies, including Replay, Elastic Weight Consolidation, Learning without Forgetting, and Synaptic Intelligence. Key contributions include: (i) introducing a new benchmark for CF in time series FL, (ii) conducting a comprehensive comparative analysis of state-of-the-art methods, and (iii) releasing a reproducible open-source framework. This work provides essential tools and insights for advancing continual learning in federated time-series forecasting systems.
format Preprint
id arxiv_https___arxiv_org_abs_2510_21491
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Benchmarking Catastrophic Forgetting Mitigation Methods in Federated Time Series Forecasting
Hallak, Khaled
Kem, Oudom
Machine Learning
Distributed, Parallel, and Cluster Computing
68T07, 68W15, 62M10
I.2.6; I.2.7; I.5.1; I.5.4
Catastrophic forgetting (CF) poses a persistent challenge in continual learning (CL), especially within federated learning (FL) environments characterized by non-i.i.d. time series data. While existing research has largely focused on classification tasks in vision domains, the regression-based forecasting setting prevalent in IoT and edge applications remains underexplored. In this paper, we present the first benchmarking framework tailored to investigate CF in federated continual time series forecasting. Using the Beijing Multi-site Air Quality dataset across 12 decentralized clients, we systematically evaluate several CF mitigation strategies, including Replay, Elastic Weight Consolidation, Learning without Forgetting, and Synaptic Intelligence. Key contributions include: (i) introducing a new benchmark for CF in time series FL, (ii) conducting a comprehensive comparative analysis of state-of-the-art methods, and (iii) releasing a reproducible open-source framework. This work provides essential tools and insights for advancing continual learning in federated time-series forecasting systems.
title Benchmarking Catastrophic Forgetting Mitigation Methods in Federated Time Series Forecasting
topic Machine Learning
Distributed, Parallel, and Cluster Computing
68T07, 68W15, 62M10
I.2.6; I.2.7; I.5.1; I.5.4
url https://arxiv.org/abs/2510.21491