The Chair of Business Analystics BAN is focused on statistical learning theory especially applied to network inference problems. Some novel features of the research carried out by BAN in the area of network inference are 1) the use of often non- commutative information structures, such as timing or the notion of neighborhood in graphs, as an underutilized degree of freedom that provides rich statistical structure about the information dynamics; and 2) the incorporation of causal (as opposed to correlation-only) relationship among network entities.
BAN’s mission is to develop theoretical models and practical algorithms to infer the statistical causal dynamics in complex networks, ranging from human brain all the way to financial networks. Learning the causal dynamics not only aids in understanding the functional map of the network but also in performing control tasks.
The BAN chair is led by Professor Negar Kiyavash.