Integral of a correspondence

From HandWiki

In mathematics, the integral of a correspondence is a generalization of the integration of single-valued functions to correspondences (i.e., set-valued functions).

The first notion of the integral of a correspondence is due to Aumann in 1965,[1] with a different approach by Debreu appearing in 1967.[2] Integrals of correspondences have applications in general equilibrium theory in mathematical economics,[3][4] random sets in probability theory,[5][6] partial identification in econometrics,[7] and fuzzy numbers in fuzzy set theory.[8]

Preliminaries

Correspondences

A correspondence φ:XY is a function φ:X𝒫(Y), where 𝒫(Y) is the power set of Y. That is, φ assigns each point xX with a set φ(x)Y.

Selections

A selection f of a correspondence φ:XY is a function f:XY such that f(x)φ(x) for every xX.

If X can be seen as a measure space (X,𝒳,μ) and Y as a Banach space (Y,||||), then one can define a measurable selection f as an 𝒳-measurable function[nb 1] f such that f(x)φ(x) for μ-almost all xX.[5][nb 2]

Definitions

The Aumann integral

Let (X,𝒳,μ) be a measure space and (Y,||||) a Banach space. If φ:XY is a correspondence, then the Aumann integral of φ is defined as

Xφdμ:={Xfdμ:f is a measurable selection of φ}

where the integrals Xfdμ are Bochner integrals.

Example: let the underlying measure space be ([0,1],,λ), and a correspondence φ:[0,1] be defined as φ(x)={2,3} for all x[0,1]. Then the Aumman integral of φ is Xφdλ=[2,3].

The Debreu integral

Debreu's approach to the integration of a correspondence is more restrictive and cumbersome, but directly yields extensions of usual theorems from the integration theory of functions to the integration of correspondences, such as Lebesgue's Dominated convergence theorem.[3] It uses Rådström's embedding theorem to identify convex and compact valued correspondences with subsets of a real Banach space, over which Bochner integration is straightforward.[2]

Let (X,𝒳,μ) be a measure space, (Y,||||) a Banach space, and 𝒦𝒫(Y) the set of all its convex and compact subsets. Let φ:X𝒦 be a convex and compact valued correspondence from X to Y. By Rådström's embedding theorem, 𝒦 can be isometrically embedded as a convex cone C in a real Banach space (𝒴,||||𝒴), in such a way that addition and multiplication by nonnegative real numbers in 𝒴 induces the corresponding operation in 𝒦.

Let φ*:XC be the "image" of φ under the embedding defined above, in the sense that φ*(x)C is the image of φ(x) under this embedding for every xX. For each pair of 𝒳-simple functions f,g:XC, define the metric Δ(f,g)=X||fg||𝒴dμ.

Then we say that φ is integrable if φ* is integrable in the following sense: there exists a sequence of 𝒳-simple functions (fn)n from X to C which are Cauchy in the metric Δ and converge in measure to φ*. In this case, we define the integral of φ* to be

Xφ*dμ:=limnXfndμC

where the integrals Xfndμ are again simply Bochner integrals in the space (𝒴,||||𝒴), and the result still belongs C since it is a convex cone. We then uniquely identify the Debreu integral of φ as[5]

Xφdμ:=E𝒦

such that E*=Xφ*dμC. Since every embedding is injective and surjective onto its image, the Debreu integral is unique and well-defined.

Notes

  1. Measurable in the sense of Bochner measurable: there exists a sequence (ψn)n of simple functions from X to Y such that limn||f(x)ψn(x)||=0 for μ-almost all xX.
  2. A stronger definition sometimes used requires f to be measurable and f(x)φ(x) for all xX. [4]

References

  1. Aumann, Robert J. (1965). "Integrals of Set-Valued Functions*". Journal of Mathematical Analysis and Applications 12: 1–12. doi:10.1016/0022-247X(65)90049-1. 
  2. 2.0 2.1 Debreu, Gérard (1967). "Integration of Correspondences". in Le Cam, Lucien; Neyman, Jerzy. Proceedings of the Fifth Berkeley Symposium on Mathematical Statistics and Probability, Volume 2: Contributions to Probability Theory, Part 1. University of California Press. pp. 351–372. ISBN 978-0520366701. 
  3. 3.0 3.1 Klein, Erwin; Thompson, Anthony C. (1984). Theory of Correspondences: Including Applications to Mathematical Economics. John Wiley & Sons. ISBN 0-471-88016-7. 
  4. 4.0 4.1 Border, Kim; Aliprantis, Charalambos D. (2006). Infinite Dimensional Analysis: A Hitchhiker's Guide (3rd ed.). Springer-Verlag. doi:10.1007/3-540-29587-9. ISBN 978-3540295860. https://link.springer.com/book/10.1007/3-540-29587-9. 
  5. 5.0 5.1 5.2 Molchanov, Ilya (2017). Theory of Random Sets. Probability Theory and Stochastic Modelling. 87 (2nd ed.). Springer-Verlag. doi:10.1007/978-1-4471-7349-6. ISBN 978-1852338923. https://link.springer.com/book/10.1007/978-1-4471-7349-6. 
  6. Molchanov, Ilya; Molinari, Francesca (2018). Random Sets in Econometrics (1st ed.). Cambridge University Press. ISBN 9781107121201. https://www.cambridge.org/core/books/random-sets-in-econometrics/7E99E73B330C0E9928A3DC9372D30616. 
  7. Molinari, Francesca (2020). "Microeconometrics with partial identification". in Durlauf, Steven; Hansen, Lars Peter; Heckman, James et al.. Handbook of Econometrics, vol. 7. Elsevier. pp. 355–486. ISBN 9780444636492. 
  8. Zimmermann, Hans-Jürgen (2011). Fuzzy Set Theory - and Its Applications (4th ed.). Springer Science + Business. ISBN 978-0-7923-7435-0.