Lifelong Learning by Adjusting Priors


Nov 03, 2017 (modified: Nov 03, 2017) ICLR 2018 Conference Blind Submission readers: everyone Show Bibtex
  • Abstract: In representational lifelong learning an agent aims to continually learn to solve novel tasks while updating its representation in light of previous tasks. Under the assumption that future tasks are `related’' to previous tasks, representations should be learned in such a way that they capture the common structure across learned tasks, while allowing the learner sufficient flexibility to adapt to novel aspects of a new task. We develop a framework for lifelong learning in deep neural networks that is based on generalization bounds, developed within the PAC-Bayes framework. Learning takes place through the construction of a distribution over networks based on the tasks seen so far, and its utilization for learning a new task. Thus, prior knowledge is incorporated through setting a history-dependent prior for novel tasks. We develop a gradient-based algorithm implementing these ideas, based on minimizing an objective function motivated by generalization bounds, and demonstrate its effectiveness through numerical examples. In addition to establishing the improved performance available through lifelong learning, we demonstrate the intuitive way by which prior information is manifested at different levels of the network.
  • TL;DR: We develop a lifelong learning approach to transfer learning based on PAC-Bayes theory, whereby priors are adjusted as new tasks are encountered thereby facilitating the learning of novel tasks.
  • Keywords: Lifelong learning, Transfer learning, PAC-Bayes theory