A Unifying Bayesian View of Continual Learning

Abstract

Some machine learning applications require continual learning - where data comes in a sequence of datasets, each is used for training and then permanently discarded. From a Bayesian perspective, continual learning seems straightforward: Given the model posterior one would simply use this as the prior for the next task. However, exact posterior evaluation is intractable with many models, especially with Bayesian neural networks (BNNs). Instead, posterior approximations are often sought. Unfortunately, when posterior approximations are used, prior-focused approaches do not succeed in evaluations designed to capture properties of realistic continual learning use cases. As an alternative to prior-focused methods, we introduce a new approximate Bayesian derivation of the continual learning loss. Our loss does not rely on the posterior from earlier tasks, and instead adapts the model itself by changing the likelihood term. We call these approaches likelihood-focused. We then combine prior- and likelihood-focused methods into one objective, tying the two views together under a single unifying framework of approximate Bayesian continual learning.

Presented at the Bayesian Deep Learning Workshop at Neural Information Processing Systems December 2018

References in corpus (6)

• Overcoming catastrophic forgetting in neural networks
• Riemannian Walk for Incremental Learning: Understanding Forgetting and Intransigence
• Continual Learning Through Synaptic Intelligence
• Variational Dropout and the Local Reparameterization Trick
• Online Structured Laplace Approximations For Overcoming Catastrophic Forgetting
• Towards Robust Evaluations of Continual Learning

Cited by in corpus (19)

• A Review of Uncertainty Quantification in Deep Learning: Techniques, Applications and Challenges
• A continual learning survey: Defying forgetting in classification tasks
• Towards Robust Evaluations of Continual Learning
• Continual Learning for Recurrent Neural Networks: an Empirical Evaluation
• CLeaR: An Adaptive Continual Learning Framework for Regression Tasks
• Orthogonal Gradient Descent for Continual Learning
• Unified Probabilistic Deep Continual Learning through Generative Replay and Open Set Recognition
• Improved Schemes for Episodic Memory-based Lifelong Learning
• Optimization and Generalization of Regularization-Based Continual Learning: a Loss Approximation Viewpoint
• Continual Learning in Recurrent Neural Networks
• Hierarchical Indian Buffet Neural Networks for Bayesian Continual Learning
• Continual Learning using a Bayesian Nonparametric Dictionary of Weight Factors
• Robust Importance Sampling for Error Estimation in the Context of Optimal Bayesian Transfer Learning
• Tractable Approximate Gaussian Inference for Bayesian Neural Networks
• Overcoming Catastrophic Forgetting by Generative Regularization
• Posterior Meta-Replay for Continual Learning
• Continual Learning from the Perspective of Compression
• Detecting and Adapting to Irregular Distribution Shifts in Bayesian Online Learning
• Subjectivity Learning Theory towards Artificial General Intelligence