{[ promptMessage ]}

Bookmark it

{[ promptMessage ]}

# HW6ans - 5 1 0 0 4 3 0 0 b =[3 2 6 6 x = A\b x = 1.0909...

This preview shows pages 1–2. Sign up to view the full content.

HW6 Problem 1 1(a):  See figure 1 (b): Vertices: (0,0), (0,2), (6/5,0), (12/11, 6/11). f(0,0) = 0 f(0,2) = -4 f(6/5,0) = -3.6 f(12/11, 6/11) = -4.363  * * objective function has the min. value at this vertex. 1(c): At vertex (12/11, 6/11) the Lagrange multipliers (lambdas) are 0.0909 and 0.6364 respectively.  At vertex (0,0) the Lagrange multipliers are negative, hence violated KKT condition 4.

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Problem 2 2(a): At vertex (12/11, 6/11) the objective function takes on its lowest value. The constraints that are active  are 5x1 + x2 <= 6 and 4x1 + 3x2 <=6. 2(b): At the chosen vertex from part a, the Lagrange multipliers are in violation of KKT condition 4. The  Lagrange multipliers are -0.107 and 0.338 respectively. They are computed by:           2     0     5     4   A =   0     2     1     3
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: 5 1 0 0 4 3 0 0 b = [3 2 6 6]' x = A\b; x = 1.0909 0.5455 -0.1074 0.3388 2(c): Yes, there is a feasible point other than a vertex where the objective function takes on an even lower value. It's computed by: 2 0 4 A= 0 2 3 4 3 0 b=[3 2 6]'; x=A\b x = 1.0200 0.6400 0.2400 This point satisfies KKT conditions 1 and 5 and the Lagrange multiplier (0.24) is positive which also satisfies KKT condition 4. It also passes KKT condition 3 and KKT condition 2 does not apply here....
View Full Document

{[ snackBarMessage ]}