Daniel Lee Bio

It is ironic that computers excel at logical reasoning that take humans many years of specialized training and education to learn, yet machines are still unable to perform simple everyday tasks that we take for granted. My group’s research focuses on learning representations that enable autonomous systems to efficiently reason about real-time behaviors in an uncertain world. In particular, much of our work has used low-dimensional representations to overcome the curse of dimensionality in perception, planning and control tasks. We use machine learning algorithms and computational neuroscience models, in addition to implementations on a variety of robotic platforms to study how to build better sensorimotor systems that can adapt and learn from experience.

Website