Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Explainability and Continual Learning meet Federated Learning at the Network Edge

Thomas Tsouparopoulos, Iordanis Koutsopoulos

TL;DR

The paper addresses the challenge of making Federated Learning at the network edge both interpretable and adaptable in the presence of non-i.i.d. data and limited resources. It proposes a multi-objective optimization framework that balances predictive accuracy with explainability via surrogate models, and extends this to FL across multiple clients. It also surveys decision-tree-based approaches for distributed explainability and develops continual learning with replay buffers for edge streams, including federated and distributed CL architectures. Together, these contributions offer a privacy-preserving, adaptive, and trustworthy roadmap for deploying explainable and lifelong learning at the edge, with potential extensions to wireless network conditions and service-level constraints.

Abstract

As edge devices become more capable and pervasive in wireless networks, there is growing interest in leveraging their collective compute power for distributed learning. However, optimizing learning at the network edge entails unique challenges, particularly when moving beyond conventional settings and objectives. While Federated Learning (FL) has emerged as a key paradigm for distributed model training, critical challenges persist. First, existing approaches often overlook the trade-off between predictive accuracy and interpretability. Second, they struggle to integrate inherently explainable models such as decision trees because their non-differentiable structure makes them not amenable to backpropagation-based training algorithms. Lastly, they lack meaningful mechanisms for continual Machine Learning (ML) model adaptation through Continual Learning (CL) in resource-limited environments. In this paper, we pave the way for a set of novel optimization problems that emerge in distributed learning at the network edge with wirelessly interconnected edge devices, and we identify key challenges and future directions. Specifically, we discuss how Multi-objective optimization (MOO) can be used to address the trade-off between predictive accuracy and explainability when using complex predictive models. Next, we discuss the implications of integrating inherently explainable tree-based models into distributed learning settings. Finally, we investigate how CL strategies can be effectively combined with FL to support adaptive, lifelong learning when limited-size buffers are used to store past data for retraining. Our approach offers a cohesive set of tools for designing privacy-preserving, adaptive, and trustworthy ML solutions tailored to the demands of edge computing and intelligent services.

Explainability and Continual Learning meet Federated Learning at the Network Edge

TL;DR

The paper addresses the challenge of making Federated Learning at the network edge both interpretable and adaptable in the presence of non-i.i.d. data and limited resources. It proposes a multi-objective optimization framework that balances predictive accuracy with explainability via surrogate models, and extends this to FL across multiple clients. It also surveys decision-tree-based approaches for distributed explainability and develops continual learning with replay buffers for edge streams, including federated and distributed CL architectures. Together, these contributions offer a privacy-preserving, adaptive, and trustworthy roadmap for deploying explainable and lifelong learning at the edge, with potential extensions to wireless network conditions and service-level constraints.

Abstract

As edge devices become more capable and pervasive in wireless networks, there is growing interest in leveraging their collective compute power for distributed learning. However, optimizing learning at the network edge entails unique challenges, particularly when moving beyond conventional settings and objectives. While Federated Learning (FL) has emerged as a key paradigm for distributed model training, critical challenges persist. First, existing approaches often overlook the trade-off between predictive accuracy and interpretability. Second, they struggle to integrate inherently explainable models such as decision trees because their non-differentiable structure makes them not amenable to backpropagation-based training algorithms. Lastly, they lack meaningful mechanisms for continual Machine Learning (ML) model adaptation through Continual Learning (CL) in resource-limited environments. In this paper, we pave the way for a set of novel optimization problems that emerge in distributed learning at the network edge with wirelessly interconnected edge devices, and we identify key challenges and future directions. Specifically, we discuss how Multi-objective optimization (MOO) can be used to address the trade-off between predictive accuracy and explainability when using complex predictive models. Next, we discuss the implications of integrating inherently explainable tree-based models into distributed learning settings. Finally, we investigate how CL strategies can be effectively combined with FL to support adaptive, lifelong learning when limited-size buffers are used to store past data for retraining. Our approach offers a cohesive set of tools for designing privacy-preserving, adaptive, and trustworthy ML solutions tailored to the demands of edge computing and intelligent services.

Paper Structure

This paper contains 23 sections, 14 equations, 2 figures.

Table of Contents

  1. Introduction
  2. Related Work
  3. Explainable AI with surrogate models in a Federated Learning setting
  4. A centralized approach with Multi-objective Optimization for Explainable AI
  5. Federated Multi-objective Optimization
  6. Future directions
  7. Addressing the accuracy-explainability tradeoff in an FL setting
  8. Federated Learning of personalized explainable models
  9. Decision-tree Learning
  10. Decision Tree training fundamentals
  11. Decision Tree training with Federated Learning
  12. Training a single decision tree in an FL setting
  13. Training Tree Ensembles in an FL setting
  14. Future directions
  15. Continual Learning with replay buffers
  16. ...and 8 more sections

Figures (2)

  • Figure 1: A Federated Learning (FL) setting, where explainability is attained either with a surrogate model explaining a black-box predictive model e.g. a linear regression model (upper part, (a) of the three small figures), or with a self-explainable decision tree (bottom part, (b) of the three small figures).
  • Figure 2: A Federated Continual Learning (FCL) framework wherein each client has its own buffer, and new data arrive in streams. The local models evolve by continuously adapting to new local data and feedback from the aggregator.