Meta Continual Learning

Abstract

Using neural networks in practical settings would benefit from the ability of the networks to learn new tasks throughout their lifetimes without forgetting the previous tasks. This ability is limited in the current deep neural networks by a problem called catastrophic forgetting, where training on new tasks tends to severely degrade performance on previous tasks. One way to lessen the impact of the forgetting problem is to constrain parameters that are important to previous tasks to stay close to the optimal parameters. Recently, multiple competitive approaches for computing the importance of the parameters with respect to the previous tasks have been presented. In this paper, we propose a learning to optimize algorithm for mitigating catastrophic forgetting. Instead of trying to formulate a new constraint function ourselves, we propose to train another neural network to predict parameter update steps that respect the importance of parameters to the previous tasks. In the proposed meta-training scheme, the update predictor is trained to minimize loss on a combination of current and past tasks. We show experimentally that the proposed approach works in the continual learning setting.

References in corpus (14)

• Overcoming catastrophic forgetting in neural networks
• Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
• Continual Learning Through Synaptic Intelligence
• An Empirical Investigation of Catastrophic Forgetting in Gradient-Based Neural Networks
• Gradient Episodic Memory for Continual Learning
• FearNet: Brain-Inspired Model for Incremental Learning
• Variational Continual Learning
• Continual Learning with Deep Generative Replay
• Continuous Adaptation via Meta-Learning in Nonstationary and Competitive Environments
• Learned Optimizers that Scale and Generalize
• Controlling Recurrent Neural Networks by Conceptors
• Learning Gradient Descent: Better Generalization and Longer Horizons
• Memory-based Parameter Adaptation
• Progressive Reinforcement Learning with Distillation for Multi-Skilled Motion Control

Cited by in corpus (11)

• Orthogonal Gradient Descent for Continual Learning
• Defining Benchmarks for Continual Few-Shot Learning
• Adversarial Meta-Learning
• Continual Learning with Adaptive Weights (CLAW)
• Meta-learnt priors slow down catastrophic forgetting in neural networks
• Continual Representation Learning for Biometric Identification
• Automatic Recall Machines: Internal Replay, Continual Learning and the Brain
• Understanding Continual Learning Settings with Data Distribution Drift Analysis
• Few-Shot Unsupervised Continual Learning through Meta-Examples
• A Conceptual Framework for Lifelong Learning
• Sparse Meta Networks for Sequential Adaptation and its Application to Adaptive Language Modelling