Skip to main content

Higher Information from Families of Measures

  • Conference paper
  • First Online:
Geometric Science of Information (GSI 2023)

Part of the book series: Lecture Notes in Computer Science ((LNCS,volume 14071))

Included in the following conference series:

Abstract

We define the notion of a measure family: a pre-cosheaf of finite measures over a finite set; every joint measure on a product of finite sets has an associated measure family. To each measure family there is an associated index, or “Euler characteristic”, related to the Tsallis deformation of mutual information. This index is further categorified by a (weighted) simplicial complex whose topology retains information about the correlations between various subsystems.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

Chapter
USD 29.95
Price excludes VAT (USA)
  • Available as PDF
  • Read on any device
  • Instant download
  • Own it forever
eBook
USD 69.99
Price excludes VAT (USA)
  • Available as EPUB and PDF
  • Read on any device
  • Instant download
  • Own it forever
Softcover Book
USD 89.99
Price excludes VAT (USA)
  • Compact, lightweight edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide - see info

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

Notes

  1. 1.

    See [15] for a precise categorical equivalence between commutative W\(^*\)-algebras and (localizable) measurable spaces.

  2. 2.

    The empty measure corresponds to the zero expectation value on the zero algebra.

  3. 3.

    If one works with probability measures and measure-preserving maps, \(\boxplus \) instead manifests as an operadic structure which encapsulates the ability to take convex linear combinations of probability measures; this is the approach taken by [6].

  4. 4.

    It is easy to write down a natural bijection of \([\textsf{Meas}]\) with \(\coprod _{n=0}^{\infty } (\mathbb {R}_{\ge 0})^{\times n}\), taking \(\mathbb {R}^{\times 0} {:}{=}\{\star \}\) to correspond to the empty measure. This observation can be used to equip \([\textsf{Meas}]\) with a topology as in [3].

  5. 5.

    In some sense a 2-measure is an “acyclic cosheaf” of measures.

  6. 6.

    One can define an index with respect to any cover of \(P_{\mu }\); but our primary interest will be the cover that is the complement of the finest partition of \(P_{\upmu }\).

  7. 7.

    Augmented in this context means there is an additional degree \(-1\) component and a single map from the degree 0 component to the degree \(-1\) component.

  8. 8.

    All interesting quantities are equivariant under change of total order.

  9. 9.

    If \(\mu _{P}\) does not vanish everywhere, then \(| {\texttt{S}} \mu _{\emptyset }| = |\{\star \}| = 1\). Consequently, one can show that \(\mathfrak {X}_{0}[\texttt{A}^{\boldsymbol{\mu }}] = 1 - \chi (\varDelta '_{\upmu })\).

  10. 10.

    This is a specialization of a functor from (localizable) measurable spaces to the Banach space underlying the W\(^*\)-algebra of essentially bounded measurable functions.

  11. 11.

    The classical analog of the GNS and the commutant complexes of [11] both reduce to the alternating sum of face maps complex that is used in this note.

References

  1. Stacks Project. https://stacks.math.columbia.edu. Accessed 16 May 2023

  2. Baez, J., Fritz, T., Leinster, T.: Entropy as a functor (2011). https://ncatlab.org/johnbaez/revision/Entropy+as+a+functor/55

  3. Baez, J., Fritz, T., Leinster, T.: A characterization of entropy in terms of information loss. Entropy 13(11), 1945–1957 (2011)

    Article  MathSciNet  MATH  Google Scholar 

  4. Baudot, P., Bennequin, D.: The homological nature of entropy. Entropy 17, 3253–3318 (2015). https://doi.org/10.3390/e17053253

    Article  MathSciNet  MATH  Google Scholar 

  5. Bennequin, D., Peltre, O., Sergeant-Perthuis, G., Vigneaux, J.P.: Extra-fine sheaves and interaction decompositions (2020). arXiv:2009.12646

  6. Bradley, T.D.: Entropy as a topological operad derivation. Entropy 23(9), 1195 (2021)

    Article  MathSciNet  Google Scholar 

  7. Drummond-Cole, G., Park, J.S., Terilla, J.: Homotopy probability theory I (2015). arXiv:1302.3684

  8. Drummond-Cole, G., Park, J.S., Terilla, J.: Homotopy probability theory II (2015). arXiv:1302.5385

  9. Geiko, R., Mainiero, T., Moore, G.: A categorical triality: Matrix product factors, positive maps and von Neumann bimodules (2022)

    Google Scholar 

  10. Lang, L., Baudot, P., Quax, R., Forré, P.: Information decomposition diagrams applied beyond shannon entropy: a generalization of Hu’s theorem (2022). arXiv:2202.09393

  11. Mainiero, T.: Homological tools for the quantum mechanic (2019). arXiv: 1901.02011

  12. Mainiero, T.: The secret topological life of shared information (2020). https://youtu.be/XgbZSwRlAjU, String-Math

  13. Mainiero, T.: Higher entropy. In: Symposium on Categorical Semantics of Entropy at CUNY (2022). https://youtu.be/6cFDviX0hUs,

  14. Parzygnat, A.J.: A functorial characterization of von Neumann entropy (2020). arXiv:2009.07125

  15. Pavlov, D.: Gelfand-type duality for commutative von Neumann algebras. J. Pure Appl. Algebra 226(4), 106884 (2022). arXiv:2005.05284

  16. Sergeant-Perthuis, G.: Intersection property, interaction decomposition, regionalized optimization and applications (2021). https://doi.org/10.13140/RG.2.2.19278.38729

  17. Stanley, R.P.: Combinatorics and Commutative Algebra, Progress in Mathematics, vol. 41, 2nd edn. Birkhäuser Boston Inc., Boston (1996)

    Google Scholar 

  18. Vigneaux, J.: The structure of information: from probability to homology (2017). arXiv: 1709.07807

Download references

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Tom Mainiero .

Editor information

Editors and Affiliations

Rights and permissions

Reprints and permissions

Copyright information

© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG

About this paper

Check for updates. Verify currency and authenticity via CrossMark

Cite this paper

Mainiero, T. (2023). Higher Information from Families of Measures. In: Nielsen, F., Barbaresco, F. (eds) Geometric Science of Information. GSI 2023. Lecture Notes in Computer Science, vol 14071. Springer, Cham. https://doi.org/10.1007/978-3-031-38271-0_25

Download citation

  • DOI: https://doi.org/10.1007/978-3-031-38271-0_25

  • Published:

  • Publisher Name: Springer, Cham

  • Print ISBN: 978-3-031-38270-3

  • Online ISBN: 978-3-031-38271-0

  • eBook Packages: Computer ScienceComputer Science (R0)

Publish with us

Policies and ethics