Current browse context:
stat.ML
Change to browse by:
References & Citations
Statistics > Machine Learning
Title: Dex: Incremental Learning for Complex Environments in Deep Reinforcement Learning
(Submitted on 19 Jun 2017)
Abstract: This paper introduces Dex, a reinforcement learning environment toolkit specialized for training and evaluation of continual learning methods as well as general reinforcement learning problems. We also present the novel continual learning method of incremental learning, where a challenging environment is solved using optimal weight initialization learned from first solving a similar easier environment. We show that incremental learning can produce vastly superior results than standard methods by providing a strong baseline method across ten Dex environments. We finally develop a saliency method for qualitative analysis of reinforcement learning, which shows the impact incremental learning has on network attention.
Link back to: arXiv, form interface, contact.