Learning an evolved mixture model for task-free continual learning
This addresses a more realistic and challenging setting in continual learning for AI systems that need to adapt continuously without task information.
The paper tackles the problem of task-free continual learning, where models must learn from non-stationary data streams without explicit task boundaries, by introducing an evolved mixture model that dynamically expands its architecture based on data distribution shifts. The approach achieves excellent performance, as demonstrated empirically.
Recently, continual learning (CL) has gained significant interest because it enables deep learning models to acquire new knowledge without forgetting previously learnt information. However, most existing works require knowing the task identities and boundaries, which is not realistic in a real context. In this paper, we address a more challenging and realistic setting in CL, namely the Task-Free Continual Learning (TFCL) in which a model is trained on non-stationary data streams with no explicit task information. To address TFCL, we introduce an evolved mixture model whose network architecture is dynamically expanded to adapt to the data distribution shift. We implement this expansion mechanism by evaluating the probability distance between the knowledge stored in each mixture model component and the current memory buffer using the Hilbert Schmidt Independence Criterion (HSIC). We further introduce two simple dropout mechanisms to selectively remove stored examples in order to avoid memory overload while preserving memory diversity. Empirical results demonstrate that the proposed approach achieves excellent performance.