Proof of the finite form of Jensen's inequality by induction
Proof
We argue by induction on . For each , the claim is the inequality for all and all weights with sum .
- First, a weighted average stays in . Let with , and .
- Let and . Then .
- Both and lie in , and is an interval. So . In particular, both sides of the inequality are defined.
- Base case . Then , and both sides equal .
- Induction step. Let and assume the claim for points. Let and be as in the statement.
- If , then for , since the weights are nonnegative with sum . Both sides equal .
- Otherwise . Put , and .
- The weights for are nonnegative with sum . So by step 3.
- We have , with and .
- By the definition of a convex function with , .
- By the induction hypothesis for the points , .
- Multiplying by and using step 10 gives .
- By induction, the inequality holds for every .
Idea
A weighted average of points is a two-point average of and the average of the other points. Convexity handles the two-point split, and the induction hypothesis handles . The renormalized weights are what make the inner average a genuine average.
Reading path
To understand this page from the ground up, read these 12 pages in order. The foundations come first.
Show the path as a graph
- no evidence level
- Construction of the reals from Cauchy sequences concept
- Convex function definition
- Integers object
- Mathematical induction concept
- Ordered field definition
- Rational numbers object
- Ring of integers statement
- The rationals form a field statement
- The rationals form an ordered field statement
- Order on the reals definition
- Interval definition
- Proof of the finite form of Jensen's inequality by induction proof
Where this leads
27 pages build on this one.
- Jensen's inequality
- Doob's upcrossing inequality
- Jarzynski equality
- Doob's martingale inequality
- Consistency of the maximum likelihood estimator
- Gibbs' inequality
- Atkinson index
- Risk aversion
- Log-modulus of a holomorphic function is subharmonic
- Asymptotic normality of the maximum likelihood estimator
- Wald's equation
- Optional stopping theorem
- Convex functions of subharmonic functions
- Submartingale
- Maximum entropy probability distributions
- Cross-entropy
- Crooks fluctuation theorem
- Doob's martingale convergence theorems
- Expectation–maximization algorithm
- Doob–Meyer decomposition theorem
- Hadamard three-circle theorem for subharmonic functions
- Lundberg inequality
- Gibbs–Bogoliubov inequality
- Jensen inequality as a functional inequality
- Skorokhod's embedding theorem
- Generalized mean inequality
- Doob decomposition theorem
Further reading
About this page
| Tag | 4RJNW3X9PS |
|---|---|
| URI | m2://4RJNW3X9PS |
| Kind | proof |
| Writing | W0 stub |
| MSC | 26D15 |
| Proves | Jensen's inequality |
| Uses | Convex function, Interval, Mathematical induction |
| Changed | 4 November 2026 |
Markdown · JSON · Edit on GitHub
What this page needs for W1
- an inventory row: the row of the inventory that the node accounts for (0 of 1)
- a source with a locator (0 of 1)
The writing levels say how complete the text of a page is. They do not measure its truth.
What links here
No page links to this page yet.