🌐 US-Proxy
id="topics">
TOPICS
Search

Mutual Information


The mutual information between two discrete random variables X and Y is defined to be

 I(X;Y)=sum_(x in X)sum_(y in Y)P(x,y)log_2((P(x,y))/(P(x)P(y)))
(1)

bits. Additional properties are

I(X;Y)=I(Y;X)
(2)
I(X;Y)>=0,
(3)

and

 I(X;Y)=H(X)+H(Y)-H(X,Y),
(4)

where H(X) is the Shannon entropy of the random variable X and H(X,Y) is the joint entropy of these variables.

For a joint density matrix rho_(AB), the quantum analogue is

 I(A:B)=S(rho_A)+S(rho_B)-S(rho_(AB))=D(rho_(AB)Orho_A tensor rho_B),
(5)

where S is von Neumann entropy, D is quantum relative entropy, and rho_A and rho_B are the reduced density matrices. All logarithms must use the same base. This quantity measures total correlations, including classical correlations, and is not by itself a measure of quantum entanglement (Watrous 2018).


See also

Conditional Entropy, Conditional Mutual Information, Courtade-Kumar Conjecture, Entropy Chain Rule, Joint Entropy, Landauer's Principle, Quantum Relative Entropy, Shannon Entropy, von Neumann Entropy

Portions of this entry contributed by Erik G. Miller

Explore with Wolfram|Alpha

References

Cover, T. M. and Thomas, J. A. Elements of Information Theory. New York: Wiley, pp. 18-26, 1991.Watrous, J. The Theory of Quantum Information. Cambridge, England: Cambridge University Press, 2018. https://doi.org/10.1017/9781316848142.

Referenced on Wolfram|Alpha

Mutual Information

Cite this as:

Weisstein, Eric W., with contributions by Erik G. Miller. "Mutual Information." From MathWorld--A Wolfram Resource. https://mathworld.wolfram.com/MutualInformation.html

Subject classifications