Optimizing Reusable Knowledge for Continual Learning via Metalearning

Abstract

When learning tasks over time, artificial neural networks suffer from a problem known as Catastrophic Forgetting (CF). This happens when the weights of a network are overwritten during the training of a new task causing forgetting of old information. To address this issue, we propose MetA Reusable Knowledge or MARK, a new method that fosters weight reusability instead of overwriting when learning a new task. Specifically, MARK keeps a set of shared weights among tasks. We envision these shared weights as a common Knowledge Base (KB) that is not only used to learn new tasks, but also enriched with new knowledge as the model learns new tasks. Key components behind MARK are two-fold. On the one hand, a metalearning approach provides the key mechanism to incrementally enrich the KB with new knowledge and to foster weight reusability among tasks. On the other hand, a set of trainable masks provides the key mechanism to selectively choose from the KB relevant weights to solve each task. By using MARK, we achieve state of the art results in several popular benchmarks, surpassing the best performing methods in terms of average accuracy by over 10% on the 20-Split-MiniImageNet dataset, while achieving almost zero forgetfulness using 55% of the number of parameters. Furthermore, an ablation study provides evidence that, indeed, MARK is learning reusable knowledge that is selectively used by each task.

References in corpus (17)

• Overcoming catastrophic forgetting in neural networks
• Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
• Continual Learning Through Synaptic Intelligence
• PathNet: Evolution Channels Gradient Descent in Super Neural Networks
• Efficient Lifelong Learning with A-GEM
• Learning multiple visual domains with residual adapters
• Three scenarios for continual learning
• On First-Order Meta-Learning Algorithms
• FearNet: Brain-Inspired Model for Incremental Learning
• Alleviating catastrophic forgetting using context-dependent gating and synaptic stabilization
• Overcoming Catastrophic Forgetting by Incremental Moment Matching
• Quantifying the Carbon Emissions of Machine Learning
• Selfless Sequential Learning
• Task Agnostic Continual Learning via Meta Learning
• Online Learned Continual Compression with Adaptive Quantization Modules
• Gradient Projection Memory for Continual Learning
• Continual Learning with Deep Artificial Neurons

Cited by in corpus (1)

• Recent Advances of Continual Learning in Computer Vision: An Overview