Uncertainty-guided Continual Learning with Bayesian Neural Networks

TL;DR

This paper introduces Uncertainty-guided Continual Bayesian Neural Networks (UCB), a method that uses weight uncertainty to adapt learning and mitigate forgetting in continual learning without needing task labels at test time.

Contribution

The paper proposes a novel continual learning approach leveraging Bayesian neural networks and uncertainty to dynamically adapt learning and perform weight pruning, reducing catastrophic forgetting.

Findings

UCB achieves superior or comparable performance on diverse datasets.

The method effectively mitigates catastrophic forgetting without task labels.

Uncertainty-based pruning retains task performance after model compression.

Abstract

Continual learning aims to learn new tasks without forgetting previously learned ones. This is especially challenging when one cannot access data from previous tasks and when the model has a fixed capacity. Current regularization-based continual learning algorithms need an external representation and extra computation to measure the parameters' \textit{importance}. In contrast, we propose Uncertainty-guided Continual Bayesian Neural Networks (UCB), where the learning rate adapts according to the uncertainty defined in the probability distribution of the weights in networks. Uncertainty is a natural way to identify \textit{what to remember} and \textit{what to change} as we continually learn, and thus mitigate catastrophic forgetting. We also show a variant of our model, which uses uncertainty for weight pruning and retains task performance after pruning by saving binary masks per tasks. We evaluate our UCB approach extensively on diverse object classification datasets with short and long sequences of tasks and report superior or on-par performance compared to existing approaches. Additionally, we show that our model does not necessarily need task information at test time, i.e. it does not presume knowledge of which task a sample belongs to.