Convex Functions: Constructions, Characterizations and by Jonathan M. Borwein

By Jonathan M. Borwein

Like differentiability, convexity is a typical and strong estate of services that performs an important position in lots of parts of arithmetic, either natural and utilized. It ties jointly notions from topology, algebra, geometry and research, and is a crucial software in optimization, mathematical programming and video game thought. This publication, that is the made of a collaboration of over 15 years, is exclusive in that it makes a speciality of convex features themselves, instead of on convex research. The authors discover a few of the sessions and their features and purposes, treating convex features in either Euclidean and Banach areas. The publication can both be learn sequentially for a graduate path, or dipped into through researchers and practitioners. every one bankruptcy encompasses a number of particular examples, and over six hundred workouts are integrated, ranging in hassle from early graduate to investigate point.

Show description

Read or Download Convex Functions: Constructions, Characterizations and Counterexamples PDF

Similar construction books

Extra info for Convex Functions: Constructions, Characterizations and Counterexamples

Example text

Suppose U ⊂ E is a convex set, and f : U → R is a convex function. Show that ∂f : U → 2E is a monotone mapping, that is, y∗ − x∗ , y − x ≥ 0 whenever x, y ∈ U and x∗ ∈ ∂f (x), y∗ ∈ ∂f (y). In particular, if U is open and f is Gâteaux differentiable on U , then ∇f (y) − ∇f (x), x − y ≥ 0 for all x, y ∈ U . 2 Differentiability 43 Hint. Use the subdifferential inequality to show y∗ − x∗ , y − x = y∗ ( y − x) + x∗ (x − y) ≥ f ( y) − f (x) + f (x) − f ( y) = 0. 22 (Continuity properties of subdifferentials).

Then the subdifferential inequality implies a + φ(u) ≤ f (u) for all u ∈ U , and so f (λx + (1 − λ)y) = φ(λx + (1 − λ)y) + a = λ(φ(x) + a) + (1 − λ)(φ(y) + a) ≤ λf (x) + (1 − λ)f ( y) as desired. An elementary relationship between subgradients and directional derivatives is recorded as follows. 16. Suppose φ ∈ ∂f (¯x). Then φ, d ≤ f (¯x; d) whenever the right-hand side is defined. Proof. This follows by taking the limit as t → 0+ in the following inequality. φ(d) = φ(td) f (t x¯ + d) − f (¯x) ≤ .

Given a nonempty set M ⊂ E, the core of M is defined by x ∈ core M if for each h ∈ SE , there exists δ > 0 so that x + th ∈ M for all 0 ≤ t ≤ δ. It is clear from the definition that, int M ⊂ core M . 13. Suppose C ⊂ E is convex. Then x0 ∈ core C if and only if x0 ∈ int C. 4). Proof. 7. The utility of the core arises in the convex context because it is often easier to check than interior, and it is naturally suited for studying directionally defined concepts, such as the directional derivative which now introduce.

Download PDF sample

Rated 4.27 of 5 – based on 35 votes