Abstract
Divergence functions play a central role in information geometry. Given a manifold \(\mathfrak {M}\), a divergence function\(\mathcal {D}\) is a smooth, nonnegative function on the product manifold \(\mathfrak {M}\times \mathfrak {M}\) that achieves its global minimum of zero (with semi-positive definite Hessian) at those points that form its diagonal submanifold \(\varDelta _{\mathfrak {M}} \subset \mathfrak {M}\times \mathfrak {M}\). In this chapter, we review how such divergence functions induce (i) a statistical structure (i.e., a Riemannian metric with a pair of conjugate affine connections) on \(\mathfrak {M}\); (ii) a symplectic structure on \(\mathfrak {M}\times \mathfrak {M}\) if they are “proper”; (iii) a Kähler structure on \(\mathfrak {M}\times \mathfrak {M}\) if they further satisfy a certain condition. It is then shown that the class of \(\mathcal {D}_\varPhi \)-divergence functions [23], as induced by a strictly convex function\(\varPhi \) on \(\mathfrak {M}\), satisfies all these requirements and hence makes \(\mathfrak {M}\times \mathfrak {M}\) a Kähler manifold (with Kähler potential given by \(\varPhi \)). This provides a larger context for the \(\alpha \)-Hessian structure induced by the \(\mathcal {D}_\varPhi \)-divergence on \(\mathfrak {M}\), which is shown to be equiaffine admitting \(\alpha \)-parallel volume forms and biorthogonal coordinates generated by \(\varPhi \) and its convex conjugate \(\varPhi ^{*}\). As the \(\alpha \)-Hessian structure is dually flat for \(\alpha = \pm 1\), the \(\mathcal {D}_\varPhi \)-divergence provides richer geometric structures (compared to Bregman divergence) to the manifold \(\mathfrak {M}\) on which it is defined.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
A holonomic coordinate system means that the coordinates have been properly “scaled” in unit-length with respect to each other such that the directional derivatives commute: their Lie bracket \([\partial _i, \partial _j] = \partial _i \partial _j - \partial _j \partial _i = 0\), i.e., the mixed partial derivatives are exchangeable in their order of application.
- 2.
This component-wise notation of Riemann curvature tensor followed standard differential geometry textbook, such as [16]. On the other hand, information geometers, such as [2], adopt the notation that \(R(\partial _i, \partial _j) \partial _k = \sum _{l} R^{l}_{ijk} \partial _l\), with \(R_{ijkl} = \sum _{l} R^{m}_{ijk} g_{ml}\).
- 3.
- 4.
The functional argument of \(x\) (or \(u\)-below) indicates that \(x\)-coordinate system (or \(u\)-coordinate system) is being used. Recall from Sect. 1.2.5 that under \(x\) (\(u\), resp) local coordinates, \(g\) and \(\varGamma \), in component forms, are expressed by lower (upper, resp) indices.
References
Amari, S.: Differential Geometric Methods in Statistics. Lecture Notes in Statistics, vol. 28. Springer, New York (1985) (Reprinted in 1990)
Amari, S., Nagaoka, H.: Method of Information Geometry. AMS Monograph. Oxford University Press, Oxford (2000)
Barndorff-Nielsen, O.E., Jupp, P.E.: Yorks and symplectic structures. J. Stat. Plan. Inference 63, 133–146 (1997)
Bregman, L.M.: The relaxation method of finding the common point of convex sets and its application to the solution of problems in convex programming. USSR Comput. Math. Phys. 7, 200–217 (1967)
Calin, O., Matsuzoe, H., Zhang. J.: Generalizations of conjugate connections. In: Sekigawa, K., Gerdjikov, V., Dimiev, S. (eds.) Trends in Differential Geometry, Complex Analysis and Mathematical Physics: Proceedings of the 9th International Workshop on Complex Structures and Vector Fields, pp. 24–34. World Scientific Publishing, Singapore (2009)
Csiszár, I.: On topical properties of f-divergence. Studia Mathematicarum Hungarica 2, 329–339 (1967)
Dombrowski, P.: On the geometry of the tangent bundle. Journal fr der reine und angewandte Mathematik 210, 73–88 (1962)
Eguchi, S.: Second order efficiency of minimum contrast estimators in a curved exponential family. Ann. Stat. 11, 793–803 (1983)
Eguchi, S.: Geometry of minimum contrast. Hiroshima Math. J. 22, 631–647 (1992)
Lauritzen, S.: Statistical manifolds. In: Amari, S., Barndorff-Nielsen, O., Kass, R., Lauritzen, S., Rao, C.R. (eds.) Differential Geometry in Statistical Inference. IMS Lecture Notes, vol. 10, pp. 163–216. Institute of Mathematical Statistics, Hayward (1987)
Matsuzoe, H.: On realization of conformally-projectively flat statistical manifolds and the divergences. Hokkaido Math. J. 27, 409–421 (1998)
Matsuzoe, H., Inoguchi, J.: Statistical structures on tangent bundles. Appl. Sci. 5, 55–65 (2003)
Matsuzoe, H., Takeuchi, J., Amari, S.: Equiaffine structures on statistical manifolds and Bayesian statistics. Differ. Geom. Appl. 24, 567–578 (2006)
Matsuzoe, H.: Computational geometry from the viewpoint of affine differential geometry. In: Nielsen, F. (ed.) Emerging Trends in Visual Computing, pp. 103–123. Springer, Berlin, Heidelberg (2009)
Matsuzoe, M.: Statistical manifolds and affine differential geometry. Adv. Stud. Pure Math. 57, 303–321 (2010)
Nomizu, K., Sasaki, T.: Affine Differential Geometry—Geometry of Affine Immersions. Cambridge University Press, Cambridge (1994)
Ohara, A., Matsuzoe, H., Amari, S.: Conformal geometry of escort probability and its applications. Mod. Phys. Lett. B 26, 1250063 (2012)
Shima, H.: Hessian Geometry. Shokabo, Tokyo (2001) (in Japanese)
Shima, H., Yagi, K.: Geometry of Hessian manifolds. Differ. Geom. Appl. 7, 277–290 (1997)
Simon, U.: Affine differential geometry. In: Dillen, F., Verstraelen, L. (eds.) Handbook of Differential Geometry, vol. I, pp. 905–961. Elsevier Science, Amsterdam (2000)
Uohashi, K.: On \(\alpha \)-conformal equivalence of statistical manifolds. J. Geom. 75, 179–184 (2002)
Yano, K., Ishihara, S.: Tangent and Cotangent Bundles: Differential Geometry, vol. 16. Dekker, New York (1973)
Zhang, J.: Divergence function, duality, and convex analysis. Neural Comput. 16, 159–195 (2004)
Zhang, J.: Referential duality and representational duality on statistical manifolds. Proceedings of the 2nd International Symposium on Information Geometry and Its Applications, Tokyo, pp. 58–67 (2006)
Zhang, J.: A note on curvature of \(\alpha \)-connections on a statistical manifold. Ann. Inst. Stat. Math. 59, 161–170 (2007)
Zhang, J., Matsuzoe, H.: Dualistic differential geometry associated with a convex function. In: Gao, D.Y., Sherali, H.D. (eds.) Advances in Applied Mathematics and Global Optimization (Dedicated to Gilbert Strang on the occasion of his 70th birthday), Advances in Mechanics and Mathematics, vol. III, Chap. 13, pp. 439–466. Springer, New York (2009)
Zhang, J.: Nonparametric information geometry: From divergence function to referential-representational biduality on statistical manifolds. Entropy 15, 5384–5418 (2013)
Zhang, J., Li, F.: Symplectic and Kähler structures on statistical manifolds induced from divergence functions. In: Nielson, F., Barbaresco, F. (eds.) Proceedings of the 1st International Conference on Geometric Science of Information (GSI2013), pp. 595–603 (2013)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2014 Springer International Publishing Switzerland
About this chapter
Cite this chapter
Zhang, J. (2014). Divergence Functions and Geometric Structures They Induce on a Manifold. In: Nielsen, F. (eds) Geometric Theory of Information. Signals and Communication Technology. Springer, Cham. https://doi.org/10.1007/978-3-319-05317-2_1
Download citation
DOI: https://doi.org/10.1007/978-3-319-05317-2_1
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-05316-5
Online ISBN: 978-3-319-05317-2
eBook Packages: EngineeringEngineering (R0)