The lecture by Hiroshi Matusuzoe from Nagoya Institute of Technology will take place at the Complexity Science Hub Vienna.
If you are interested in participating, please email to firstname.lastname@example.org
A deformed exponential family is a natural generalization of an ordinary exponential family, and it is useful for robust statistics. However, the standard maximum likelihood method does not work efficiently for deformed exponential families.
In this talk, we introduce generalized estimating functions for deformed exponential families, and give information geometric interpretations for such estimating functions. In particular, we construct dually flat structures and divergence functions on deformed exponential families from the viewpoint of unbiasedness of estimating functions.
Hiroshi Matusuzoe graduated from the Graduate School of Information Sciences, Tohoku University, in 1999. Now he is a professor at Nagoya Institute of Technology. His research activities are in the area of differential geometry and differential geometrical method of mathematical science. He holds a doctoral degree of information science.