Concentration of measure
In
The concentration of measure phenomenon was put forth in the early 1970s by
, and others.The general setting
Let be a metric space with a measure on the Borel sets with . Let
where
is the -extension (also called -fattening in the context of the Hausdorff distance) of a set .
The function is called the concentration rate of the space . The following equivalent definition has many applications:
where the supremum is over all 1-Lipschitz functions , and the median (or Levy mean) is defined by the inequalities
Informally, the space exhibits a concentration phenomenon if decays very fast as grows. More formally, a family of metric measure spaces is called a Lévy family if the corresponding concentration rates satisfy
and a normal Lévy family if
for some constants . For examples see below.
Concentration on the sphere
The first example goes back to
for suitable , has the smallest -extension (for any ).
Applying this to sets of measure (where ), one can deduce the following concentration inequality:
- ,
where are universal constants. Therefore meet the definition above of a normal Lévy family.
Vitali Milman applied this fact to several problems in the local theory of Banach spaces, in particular, to give a new proof of Dvoretzky's theorem.
Concentration of measure in physics
All classical statistical physics is based on the concentration of measure phenomena: The fundamental idea (‘theorem’) about equivalence of ensembles in thermodynamic limit (
When the number of particles is large, then the difference between average values of the macroscopic variables for the canonical and microcanonical ensembles tends to zero, and their fluctuations are explicitly evaluated. These results are proven rigorously under some regularity conditions on the energy function E by Khinchin (1943).[8] The simplest particular case when E is a sum of squares was well-known in detail before Khinchin and Lévy and even before Gibbs and Einstein. This is the Maxwell–Boltzmann distribution of the particle energy in ideal gas.
The microcanonical ensemble is very natural from the naïve physical point of view: this is just a natural equidistribution on the isoenergetic hypersurface. The canonical ensemble is very useful because of an important property: if a system consists of two non-interacting subsystems, i.e. if the energy E is the sum, , where are the states of the subsystems, then the equilibrium states of subsystems are independent, the equilibrium distribution of the system is the product of equilibrium distributions of the subsystems with the same T. The equivalence of these ensembles is the cornerstone of the mechanical foundations of thermodynamics.
Other examples
- Borell–TIS inequality
- Gaussian isoperimetric inequality
- McDiarmid's inequality
- Talagrand's concentration inequality
- Asymptotic equipartition property
References
- .
- ^ "The concentration of , ubiquitous in the probability theory and statistical mechanics, was brought to geometry (starting from Banach spaces) by Vitali Milman, following the earlier work by Paul Lévy" - M. Gromov, Spaces and questions, GAFA 2000 (Tel Aviv, 1999), Geom. Funct. Anal. 2000, Special Volume, Part I, 118–161.
- ^ "The idea of concentration of measure (which was discovered by V.Milman) is arguably one of the great ideas of analysis in our times. While its impact on Probability is only a small part of the whole picture, this impact should not be ignored." - M. Talagrand, A new look at independence, Ann. Probab. 24 (1996), no. 1, 1–34.
- ^ Gibbs, Josiah Willard (1902). Elementary Principles in Statistical Mechanics (PDF). New York, NY: Charles Scribner's Sons.
- . Retrieved January 21, 2020.
- ^ Einstein, Albert (1904). "Eine Theorie der Grundlagen der Thermodynamik [A Theory of the Foundations of Thermodynamics]" (PDF). Annalen der Physik. Series 4. 11: 417–433. Retrieved January 21, 2020.
- . Retrieved January 21, 2020.
- ^ Khinchin, Aleksandr Y. (1949). Mathematical foundations of statistical mechanics [English translation from the Russian edition, Moscow, Leningrad, 1943]. New York, NY: Courier Corporation. Retrieved January 21, 2020.
Further reading
- Ledoux, Michel (2001). The Concentration of Measure Phenomenon. American Mathematical Society. ISBN 0-8218-2864-9.
- Giannopoulos, A. A.; .
External links
- Huszár, Ferenc (November 9, 2017). "Gaussian Distributions are Soap Bubbles". – blog post illustrating one of the implications of concentration of measure