Binary entropy function

In information theory, the binary entropy function, denoted $\operatorname {H} (p)$ or $\operatorname {H} _{\text{b}}(p)$ , is defined as the entropy of a Bernoulli process with probability $p$ of one of two values. It is a special case of $\mathrm {H} (X)$ , the entropy function. Mathematically, the Bernoulli trial is modelled as a random variable $X$ that can take on only two values: 0 and 1, which are mutually exclusive and exhaustive.

If $\operatorname {Pr} (X=1)=p$ , then $\operatorname {Pr} (X=0)=1-p$ and the entropy of $X$ (in shannons) is given by

\operatorname {H} (X)=\operatorname {H} _{\text{b}}(p)=-p\log _{2}p-(1-p)\log _{2}(1-p)

,

where $0\log _{2}0$ is taken to be 0. The logarithms in this formula are usually taken (as shown in the graph) to the base 2. See binary logarithm.

When $p={\tfrac {1}{2}}$ , the binary entropy function attains its maximum value. This is the case of an unbiased coin flip.

$\operatorname {H} (p)$ is distinguished from the entropy function $\mathrm {H} (X)$ in that the former takes a single real number as a parameter whereas the latter takes a distribution or random variable as a parameter. Sometimes the binary entropy function is also written as $\operatorname {H} _{2}(p)$ . However, it is different from and should not be confused with the Rényi entropy, which is denoted as $\mathrm {H} _{2}(X)$ .

Explanation

In terms of information theory, entropy is considered to be a measure of the uncertainty in a message. To put it intuitively, suppose $p=0$ . At this probability, the event is certain never to occur, and so there is no uncertainty at all, leading to an entropy of 0. If $p=1$ , the result is again certain, so the entropy is 0 here as well. When $p=1/2$ , the uncertainty is at a maximum; if one were to place a fair bet on the outcome in this case, there is no advantage to be gained with prior knowledge of the probabilities. In this case, the entropy is maximum at a value of 1 bit. Intermediate values fall between these cases; for instance, if $p=1/4$ , there is still a measure of uncertainty on the outcome, but one can still predict the outcome correctly more often than not, so the uncertainty measure, or entropy, is less than 1 full bit.

Properties

Derivative

The derivative of the binary entropy function may be expressed as the negative of the logit function:

{d \over dp}\operatorname {H} _{\text{b}}(p)=-\operatorname {logit} _{2}(p)=-\log _{2}\left({\frac {p}{1-p}}\right)

.

{d^{2} \over dp^{2}}\operatorname {H} _{\text{b}}(p)=-{\frac {1}{p(1-p)\ln 2}}

Convex conjugate

The convex conjugate (specifically, the Legendre transform) of the binary entropy (with base e) is the negative softplus function. This is because (following the definition of the Legendre transform: the derivatives are inverse functions) the derivative of negative binary entropy is the logit, whose inverse function is the logistic function, which is the derivative of softplus.

Softplus can be interpreted as logistic loss, so by duality, minimizing logistic loss corresponds to maximizing entropy. This justifies the principle of maximum entropy as loss minimization.

Taylor series

The Taylor series of the binary entropy function at 1/2 is

\operatorname {H} _{\text{b}}(p)=1-{\frac {1}{2\ln 2}}\sum _{n=1}^{\infty }{\frac {(1-2p)^{2n}}{n(2n-1)}}

which converges to the binary entropy function for all values $0\leq p\leq 1$ .

Bounds

The following bounds hold for $0<p<1$ :^[1]

\ln(2)\cdot \log _{2}(p)\cdot \log _{2}(1-p)\leq H_{\text{b}}(p)\leq \log _{2}(p)\cdot \log _{2}(1-p)

and

4p(1-p)\leq H_{\text{b}}(p)\leq (4p(1-p))^{(1/\ln 4)}

where $\ln$ denotes natural logarithm.

References

^ Topsøe, Flemming (2001). "Bounds for entropy and divergence for distributions over a two-element set". JIPAM. Journal of Inequalities in Pure & Applied Mathematics. 2 (2): Paper No. 25, 13 p.-Paper No. 25, 13 p.