Abstract
It is shown that the l 1-distance in the space of probability assignments on a finite set Ω provides a criterion to judge whether two assignments are too close to each other to be distinguished by a statistical test. The criterion is independent of the number of elements of Ω. Other notions of distance are also discussed.
References
Abe, S.: Phys. Rev. E 66, 046134 (2002)
Souza, A.M.C., Tsallis, C.: Phys. Lett. A 319, 273 (2003)
Lesche, B.: Phys. Rev. E 70, 017102 (2004)
Naudts, J.: Rev. Math. Phys. 16, 809 (2004)
Curado, E.M.F., Nobre, F.D.: Phys. A 335, 94 (2004)
Kaniadakis, G., Scarfone, A.M.: Phys. A 340, 102 (2004)
Abe, A., Kaniadakis, G., Scarfone, A.M.: J. Phys. A: Math. Gen. 37, 10513 (2004)
Abe, S.: Phys. Rev. E 72, 028102 (2005)
Ruch, E.: Theor. Chim. Acta 38, 167 (1975)
Lesche, B.: J. Stat. Phys. 27, 419 (1982)
Reif, F.: Fundamentals of Statistical and Thermal Physics. McGraw–Hill, New York (1965)
Lindgren, B.W.: Statistical Theory, 4th edn. Chapman & Hall, New York (1993)
Csiszár, I.: Stud. Sci. Math. Hung. 2, 299 (1967)
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Abe, S., Lesche, B. & Mund, J. How Should the Distance of Probability Assignments Be Judged?. J Stat Phys 128, 1189–1196 (2007). https://doi.org/10.1007/s10955-007-9344-7
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10955-007-9344-7