My primary research goal is to develop techniques for adaptive autonomous agents learning on streams of data. My research focus to achieve this goal is on reinforcement learning and representation learning. In particular, I care about efficient, practical algorithms that enable learning from large amounts of data.

So far, I have focused on principled optimization approaches for representation learning, particularly looking at sparse representations and recurrent architectures for partially observable domains. I have also been working on off-policy reinforcement learning, which enables learning about many different policies in parallel from a single stream of interaction with the environment. My life goal is to make advances in representation learning for reinforcement learning, which I believe is the one of the biggest scientific hurdles for AI and autonomous agents.