Exploring the potential of Direct Feedback Alignment for Continual Learning

TMLR Paper2975 Authors

Abstract:

Real-world applications of machine learning require robustness to shifts in the data distribution over time. A critical limitation of standard artificial neural networks trained with backpropagation (BP) is their susceptibility to catastrophic forgetting: they “forget” prior knowledge when trained on a new task, while biological neural networks tend to be more robust to catastrophic forgetting. While various algorithmic ways of mitigating
catastrophic forgetting have been proposed, developing an algorithm that is capable of learning continuously remains an open problem. Motivated by recent theoretical results, here we explore whether a biologically inspired learning algorithm like Direct feedback Alignment (DFA) can mitigate catastrophic forgetting in artificial neural networks. We train fully-connected networks on several continual learning benchmarks using DFA and compare its performance to vanilla back propagation, random features, and other continual learning algorithms. We find that an inherent bias of DFA, called “degeneracy breaking”, leads to low average forgetting on common continual learning benchmarks when using DFA in the Domain-Incremental learning scenario and in the Task-Incremental learning scenario. We show how to control the trade-off between learning and forgetting with DFA, and relate different modes of using DFA to other methods in the field.

Submission Length: Regular submission (no more than 12 pages of main content)

Assigned Action Editor: ~Laurent_Charlin1

Submission Number: 2975