Lifelong Learning using Eigentasks: Task Separation, Skill Acquisition, and Selective Transfer

Abstract

We introduce the eigentask framework for lifelong learning. An eigentask is a pairing of a skill that solves a set of related tasks, paired with a generative model that can sample from the skill's input space. The framework extends generative replay approaches, which have mainly been used to avoid catastrophic forgetting, to also address other lifelong learning goals such as forward knowledge transfer. We propose a wake-sleep cycle of alternating task learning and knowledge consolidation for learning in our framework, and instantiate it for lifelong supervised learning and lifelong RL. We achieve improved performance over the state-of-the-art in supervised continual learning, and show evidence of forward knowledge transfer in a lifelong RL application in the game Starcraft2.

Accepted at the 4th Lifelong Machine Learning Workshop at the Thirty-seventh International Conference on Machine Learning (ICML) 2020

References in corpus (17)

• Auto-Encoding Variational Bayes
• How transferable are features in deep neural networks?
• Continual Learning Through Synaptic Intelligence
• StarCraft II: A New Challenge for Reinforcement Learning
• IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures
• Three scenarios for continual learning
• Deep Unsupervised Clustering with Gaussian Mixture Variational Autoencoders
• Alleviating catastrophic forgetting using context-dependent gating and synaptic stabilization
• Generative replay with feedback connections as a general strategy for continual learning
• A Deep Hierarchical Approach to Lifelong Learning in Minecraft
• Lifelong Generative Modeling
• On Quadratic Penalties in Elastic Weight Consolidation
• Continual Unsupervised Representation Learning
• Learn to Grow: A Continual Structure Learning Framework for Overcoming Catastrophic Forgetting
• Life-Long Disentangled Representation Learning with Cross-Domain Latent Homologies
• Deep Unsupervised Clustering Using Mixture of Autoencoders
• Multi-Task Incremental Learning for Object Detection