Convexity
Let be a convex set. A function is convex if
whenever and
In geometric terms, this means a @chord joining any two points on the graph of lies on or above the graph between them.
Referenced by (13 direct, 8 transitive)
Direct references:
- Log is Concave
- Strictly Convex Function
- Concave Function
- Adding an Affine Function Preserves Convexity
- Affine Precomposition Preserves Convexity
- Second Derivative Test for Convexity
- Convexity of Relative Entropy
- convexity-of-relative-entropy-intuition
- Concavity and Convexity of Mutual Information
- proof-of-concavity-and-convexity-of-mutual-information
- Jensen's Inequality
- proof-of-jensens-inequality
- proof-of-probability-vectors-form-a-convex-compact-set
Starting with convex function, is strictly convex if the inequality is strict whenever
A function is said to be concave if is a convex function.
Referenced by (13 direct, 2 transitive)
Direct references:
- Log is Concave
- proof-of-log-is-concave
- Adding an Affine Function Preserves Convexity
- Affine Precomposition Preserves Convexity
- Second Derivative Test for Convexity
- proof-of-gibbs-inequality
- proof-of-uniform-distribution-maximizes-entropy
- Concavity of Entropy
- proof-of-concavity-of-entropy
- Concavity and Convexity of Mutual Information
- proof-of-concavity-and-convexity-of-mutual-information
- Jensen's Inequality
- note-7
Let be a convex set, let and let be affine. If is convex, then and are convex. If is concave, then and are concave. A linear function is affine, so this covers adding or subtracting a linear function.
TODO: prove this.
Referenced by (1 direct)
Direct references:
Let and be convex sets, let be affine, and let If is convex, then is convex on If is concave, then is concave on
TODO: prove this.
Referenced by (1 direct)
Direct references:
Let be twice differentiable on an open interval Then is convex on if and only if for all and concave if and only if for all
Referenced by (1 direct)
Direct references:
A convex combination is a linear combination of points where all @coefficients are non-negative and sum to 1.
Every convex combination of two points lies on the @line-segment between the points.