Philipp Becker

  • Adenauerring 4

    Gebäude 50.21

    76131 Karlsruhe

About me

I joined the Autonomous Learning Robot group in January 2020 to continue my PhD, which I started in May 2019 at the Bosch Center for Artificial Intelligence and the University of Tübingen. Before that, I received a Bachelor's in Computer Science and a Master's in Autonomous Systems from TU Darmstadt. My research focuses mainly on probabilistic modeling for Model-Based Reinforcement Learning, general time-series, and multi-modal data in the context of robotics. I already focused on those topics during my Bachelor's and Master's theses. Time series are ubiquitous in robotics, and acting based on information received over time is crucial for many real-world tasks. To this end, we must accumulate partial information over time and appropriately represent uncertainties to allow safe and optimal behavior. As, in the real world, there is usually more than one possible way to solve a given problem, multi-modal distributions are essential in modeling safe and versatile behavior - in both robots and humans. I am also interested in general reinforcement learning.