Model-Free Generative Replay for Lifelong Reinforcement Learning: Application to Starcraft-2

Zachary Alan Daniels, Aswin Raghavan, Jesse Hostetler, Abrar Rahman, Indranil Sur, Michael Piacentino, Ajay Divakaran, Roberto Corizzo, Kamil Faber, Nathalie Japkowicz, Michael Baron, James Seale Smith, Sahana Pramod Joshi, Zsolt Kira, Cameron Ethan Taylor, Mustafa Burak Gurbuz, Constantine Dovrolis, Tyler L. Hayes, Christopher Kanan, Jhair Gallardo

Abstract: One approach to meet the challenges of deep lifelong reinforcement learning (LRL) is careful management of the agent’s learning experiences, in order to learn (without forgetting) and build interna...