Continual Learning with Node-Importance based Adaptive Group Sparse Regularization

Abstract

We propose a novel regularization-based continual learning method, dubbed as Adaptive Group Sparsity based Continual Learning (AGS-CL), using two group sparsity-based penalties. Our method selectively employs the two penalties when learning each node based its the importance, which is adaptively updated after learning each new task. By utilizing the proximal gradient descent method for learning, the exact sparsity and freezing of the model is guaranteed, and thus, the learner can explicitly control the model capacity as the learning continues. Furthermore, as a critical detail, we re-initialize the weights associated with unimportant nodes after learning each task in order to prevent the negative transfer that causes the catastrophic forgetting and facilitate efficient learning of new tasks. Throughout the extensive experimental results, we show that our AGS-CL uses much less additional memory space for storing the regularization parameters, and it significantly outperforms several state-of-the-art baselines on representative continual learning benchmarks for both supervised and reinforcement learning tasks.

References in corpus (10)

• Playing Atari with Deep Reinforcement Learning
• Riemannian Walk for Incremental Learning: Understanding Forgetting and Intransigence
• Continual Learning Through Synaptic Intelligence
• Learning Structured Sparsity in Deep Neural Networks
• FearNet: Brain-Inspired Model for Incremental Learning
• Variational Continual Learning
• Overcoming catastrophic forgetting with hard attention to the task
• Continual Learning with Deep Generative Replay
• Discrimination-aware Channel Pruning for Deep Neural Networks
• Continual Learning via Neural Pruning

Cited by in corpus (10)

• Continual Learning in Sensor-based Human Activity Recognition: an Empirical Benchmark Analysis
• Recent Advances of Continual Learning in Computer Vision: An Overview
• CPR: Classifier-Projection Regularization for Continual Learning
• Preserving Earlier Knowledge in Continual Learning with the Help of All Previous Feature Extractors
• Generalisation Guarantees for Continual Learning with Orthogonal Gradient Descent
• Formalizing the Generalization-Forgetting Trade-off in Continual Learning
• Graceful Degradation and Related Fields
• Understanding Catastrophic Forgetting and Remembering in Continual Learning with Optimal Relevance Mapping
• Group and Exclusive Sparse Regularization-based Continual Learning of CNNs
• Discriminative Distillation to Reduce Class Confusion in Continual Learning