Deep Class Conditional Gaussians for Continual Learning

Thomas L Lee, Amos Storkey

Keywords: Contiual Learning, Lifelong Learning, Bayesian, Emprical Bayes, Probabilistic Machine Learning

Abstract: The current state of the art for continual learning with frozen, pre-trained embedding networks are simple probabilistic models defined over the embedding space, for example class conditional Gaussians. However, as of yet, in the task-incremental online setting, it has been an open question how to extend these methods to when the embedding function has to be learned from scratch. In this paper, we propose an empirical Bayesian framework that works by storing a fixed number of examples in memory which are used to calculate the posterior of the probabilistic model and a conditional marginal likelihood term used to fit the embedding function. The learning of the embedding function can be interpreted as using a variant of experience replay, which is a highly performative method for continual learning. As part of our framework, we decide which examples to store by selecting the subset that minimises the KL divergence between the true posterior and the posterior induced by the subset, which is shown to be necessary to achieve good performance. We demonstrate the performance of our method on a range of task-incremental online settings, including those with overlapping tasks which thus far have been under-explored. Our method outperforms all other methods, including several other replay-based methods, evidencing the potential of our approach.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics

Submission Guidelines: Yes

Please Choose The Closest Area That Your Submission Falls Into: Probabilistic Methods (eg, variational inference, causal inference, Gaussian processes)

TL;DR: We present an empirical Bayesian method to solve the problem in continual learning of how to use simple metric-based probabilistic models when the embedding function must be learnt online.