Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: THEend8_
Assignment
1. Suppose A ∈ Rn ×n is an invertible matrix. Show that AT A is positive definite.
2. Consider f : R2 → R given by f(x1 , x2 ) = 2x1 x2 − x1(2) − x2(2). Compute the Hessian ∇2 f(x) and conclude if f is a convex function.
3. Show that a function f : Ω ⊆ Rn → R ∈ C1 is convex iff
f(z) ≥ f(x) + ∇f(x)T (z − x) ∀x,z ∈ Ω .
4. Consider the function f(x) = xTAx where A ∈ Rn ×n. Find ∇f(x). What is ∇f(x), if A is symmetric? Note. Do not use a pre-derived formula and show all your work.
5. Consider the Euclidean ball B = {x : x1(2) + x2(2) ≤ 1}. Find all feasible directions d at x = (0, 0) and y = (1, 0). Justify your work.
Suggested problems (not for submission)
1. Show that the intersection of two convex sets is convex.
2. Let C ⊂ R2 be the triangle with vertices (0, 0), (a,0), and (0, a), where a > 0. Prove C is convex.
3. Consider the following problem:
minimize ∥Ax − b∥2(2)
s.t. x1 + ··· + xn = 1
x1 , ··· , xn ≥ 0
Here A ∈ Rm ×n (m > n) and b ∈ Rm . Is the problem a convex optimization problem? If yes, provide a proof. If no, explain why not, giving complete explanations.
4. Show that a function f : Ω ⊆ Rn → R ∈ C2 is convex iff ∇2 f(x) is positive semidefinite for all x ∈ Ω .
5. Let f : Rn → R be a convex function defined on a convex set Ω ⊆ Rn. Show a point is a global minimizer of f over Ω iff it is a local minimizer of f.
6. Let f : Rn → R be a strictly convex function. Show that f has at most one global minimizer.
7. Let f : Rn → R be a convex function defined on a convex set Ω ⊆ Rn. Then, the set of all global minimizers of f over Ω is a convex set.