Continual Learning using a Bayesian Nonparametric Dictionary of Weight Factors
arXiv:2004.10098
Abstract
Naively trained neural networks tend to experience catastrophic forgetting in sequential task settings, where data from previous tasks are unavailable. A number of methods, using various model expansion strategies, have been proposed recently as possible solutions. However, determining how much to expand the model is left to the practitioner, and often a constant schedule is chosen for simplicity, regardless of how complex the incoming task is. Instead, we propose a principled Bayesian nonparametric approach based on the Indian Buffet Process (IBP) prior, letting the data determine how much to expand the model complexity. We pair this with a factorization of the neural network's weight matrices. Such an approach allows the number of factors of each weight matrix to scale with the complexity of the task, while the IBP prior encourages sparse weight factor selection and factor reuse, promoting positive knowledge transfer between tasks. We demonstrate the effectiveness of our method on a number of continual learning benchmarks and analyze how weight factors are allocated and reused throughout the training.
arXiv commentProceedings of the 24th International Conference on Artificial Intelligence and Statistics (AISTATS) 2021 Post-conference updates: Fixed typo in equation (11) and updated references
References in corpus (19)
- Distilling the Knowledge in a Neural Network
- Overcoming catastrophic forgetting in neural networks
- GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium
- The Concrete Distribution: A Continuous Relaxation of Discrete Random Variables
- Continual Learning Through Synaptic Intelligence
- A Simple Unified Framework for Detecting Out-of-Distribution Samples and Adversarial Attacks
- An Empirical Investigation of Catastrophic Forgetting in Gradient-Based Neural Networks
- Three scenarios for continual learning
- Re-evaluating Continual Learning Scenarios: A Categorization and Case for Strong Baselines
- Reinforced Continual Learning
- Generative replay with feedback connections as a general strategy for continual learning
- Online Structured Laplace Approximations For Overcoming Catastrophic Forgetting
- Towards Robust Evaluations of Continual Learning
- Understanding Measures of Uncertainty for Adversarial Example Detection
- Compacting, Picking and Growing for Unforgetting Continual Learning
- A Neural Dirichlet Process Mixture Model for Task-Free Continual Learning
- A Unifying Bayesian View of Continual Learning
- Generative Adversarial Network Training is a Continual Learning Problem
- Hierarchical Indian Buffet Neural Networks for Bayesian Continual Learning