This research position is embedded in a research program on developing probabilistic (Bayesian) machine learning methods for development and deployment of in-situ operating intelligent autonomous agents. Typical applications for these agents include in-situ learning of personalized hearing aid or gesture recognition algorithms or in-situ learning by adaptive robots. In order to support this research program, we develop and maintain a "probabilistic programming" toolbox (in programming language Julia,
http://julialang.org ) to support efficient implementation of Bayesian intelligent agents, see
http://forneylab.org . A crucial goal of the toolbox is to automate probabilistic reasoning by Bayesian agents through message passing algorithms on graphs. This PhD project is aimed at further development of the toolbox, both in theoretical and practical directions. Key areas of interest include Bayesian machine learning, functional and reactive programming, probabilistic programming and automatic differentiation. The position is very suited for a programmer/developer with a serious interest in learning about modern developments in Bayesian machine learning and probabilistic programming. The starting date is preferably around September or October in 2019.
Please browse our web page
http://biaslab.org for more information on our research goals.