Certified Continual Learning for Neural Network Regression

TL;DR

This paper introduces a method called certified continual learning that maintains the verified correctness of neural networks during ongoing training, effectively addressing issues like catastrophic forgetting while preserving model utility.

Contribution

It proposes a novel approach to preserve neural network correctness properties during continual learning, enhancing existing methods with certification guarantees.

Findings

The approach effectively preserves certified correctness during continual learning.

Models maintain high utility and accuracy over multiple training phases.

The method is efficient across different neural network architectures.

Abstract

On the one hand, there has been considerable progress on neural network verification in recent years, which makes certifying neural networks a possibility. On the other hand, neural networks in practice are often re-trained over time to cope with new data distribution or for solving different tasks (a.k.a. continual learning). Once re-trained, the verified correctness of the neural network is likely broken, particularly in the presence of the phenomenon known as catastrophic forgetting. In this work, we propose an approach called certified continual learning which improves existing continual learning methods by preserving, as long as possible, the established correctness properties of a verified network. Our approach is evaluated with multiple neural networks and on two different continual learning methods. The results show that our approach is efficient and the trained models preserve their certified correctness and often maintain high utility.