February 07, 2013
Collins Conference Room
Eckehard Olbrich (Max Planck Institute for Mathematics in the Sciences)
Abstract. How can the information that a set of random variables contains about another random variable be decomposed? To what extent do different subgroups provide the same, i.e. shared or redundant, information, carry unique information or interact for the emergence of synergistic information?
Recently Williams and Beer proposed such a decomposition based on natural properties for shared information. While these properties fix the structure of the decomposition, they do not uniquely specify the values of the different terms. Therefore, we investigate additional properties such as strong symmetry and left monotonicity. We find that strong symmetry is incompatible with the properties proposed by Williams and Beer. Although left monotonicity is a very natural property for an information measure it is not fulfilled by any of the proposed measures.
We also study a geometric framework for information decompositions and ask whether it is possible to represent shared information by a family of posterior distributions.
Finally, we draw connections to the notions of shared knowledge and common knowledge in game theory. While many people believe that independent variables cannot share information, we show that in game theory independent agents can have shared knowledge, but not common knowledge.
We conclude that intuition and heuristic arguments do not suffice when arguing about information. We expect that further progress requires a more precise, operational idea of what shared information should be.
Nils Bertschinger, Johannes Rauh, Eckehard Olbrich, and Jürgen Jost. "Shared Information — New Insights and Problems in Decomposing Information in Complex Systems." arXiv:1210.5902 (2012)
Williams, P., and R. Beer. "Nonnegative decomposition of multivariate information." arXiv:1004.2515v1 (2010)
Purpose: Research Collaboration
SFI Host: David Wolpert
During an SFI Community Lecture next Wednesday night in Santa Fe, MIT's Alex Pentland describes ways the mathematical analysis of social networks is fertile ground for understanding human behavior.
Register now for SFI's 2015 Short Course — Exploring Complexity in Social Systems and Economics — August 25-27 in Santa Fe.
The Santa Fe Institute again has ranked among the world's top science and technology think tanks.
SFI is playing a leading role in an international scientific conference on complex systems science to be held this fall in Tempe, Arizona.
During an SFI Community Lecture January 14 in Santa Fe, Stanford linguist Daniel Jurafsky explored the stunningly complex language of food and what it tells us about our culture and ...