hw4: GSPPN-DUMBIFIED

This commit is contained in:
Claudio Maggioni 2021-05-28 12:22:58 +02:00
parent 25fdd4e598
commit 43f54092d6
2 changed files with 91 additions and 29 deletions

View File

@ -93,7 +93,7 @@ $$f(X_1) = -1.1304 \;\; f(X_2) = -1.59219 \;\; f(X_3) = 4.64728 \;\; f(X_4) =
To find the optimal solution, we choose $(\lambda_4, X_4)$ since $f(X_4)$ is the
smallest objective
value out of all the feasible points. Therefore, the solution to the
minimization problem is:
constrained minimization problem is:
$$X \approx \begin{bmatrix}-0.966\\-0.199\\-0.166\end{bmatrix}$$
@ -102,15 +102,70 @@ $$X \approx \begin{bmatrix}-0.966\\-0.199\\-0.166\end{bmatrix}$$
## Exercise 2.1
To reformulate the problem, we first rewrite the explicit values of $G$, $c$,
$A$ and $b$:
$A$ and $b$.
$$G = 2 \cdot \begin{bmatrix}3&0&0\\2&2.5&0\\1&2&2\end{bmatrix}$$
$$c = \begin{bmatrix}-8\\-3\\-3\end{bmatrix}$$
$$A = \begin{bmatrix}1&0&1\\0&1&1\end{bmatrix}$$
$$b = \begin{bmatrix}3\\0\end{bmatrix}$$
We first define matrix $G$ as a set of 9 unknown variables and $c$ a set of 3
unknown variables:
Then, using these variable values and the formulation given on the assignment
sheet the problem is restated in this new form.
$$G = \begin{bmatrix}g_{11}&g_{12}&g_{13}\\g_{21}&g_{22}&g_{23}\\
g_{31}&g_{32}&g_{33}\end{bmatrix} \;\;\; c =
\begin{bmatrix}c_1\\c_2\\c_3\end{bmatrix}$$
We then define $f(x)$ in the following way:
$$f(x) = \frac12 \cdot \begin{bmatrix}x_1&x_2&x_3\end{bmatrix}
\begin{bmatrix}g_{11}&g_{12}&g_{13}\\g_{21}&g_{22}&g_{23}\\
g_{31}&g_{32}&g_{33}\end{bmatrix} \begin{bmatrix}x_1\\x_2\\x_3\end{bmatrix}
- \begin{bmatrix}c_1&c_2&c_3\end{bmatrix}\begin{bmatrix}x_1\\x_2\\x_3\end{bmatrix}
=$$$$=x_1^2 \cdot \frac{g_{11}}{2} + x_2^2 \cdot \frac{g_{22}}{2} +
x_3^2 \cdot \frac{g_{33}}{2} +
\left(\frac{g_{12} + g_{21}}{2}\right) x_1 x_2
+ \left(\frac{g_{13} + g_{31}}{2}\right) x_1 x_3 +
\left(\frac{g_{23} + g_{32}}{2}\right) x_2 x_3 + c_1 x_1 + c_2 x_2 + c_3 x_3$$
Then, we equal this polynomial to the given one, finding the following values
and constraints for the coefficients of $G$ and $g$:
$$\begin{cases}g_{11} = 3 \cdot 2 = 6 \\
g_{22} = 2.5\cdot 2 = 5 \\
g_{33} = 2 \cdot 2 = 4 \\
c_1 = -8 \\
c_2 = -3 \\
c_3 = -3 \\
g_{13} + g_{31} = 1 \cdot 2 = 2 \\
g_{12} + g_{21} = 2 \cdot 2 = 4 \\
g_{23} + g_{32} = 2 \cdot 2 = 4 \end{cases}$$
As it can be seen by the system of equations above, we have infinite possibility
for choosing the components of the $G$ matrix that are not on the main diagonal.
Due to personal taste, we choose those components in such a way that the
resulting $G$ matrix is symmetric. We therefore obtain:
$$G = \begin{bmatrix}6&2&1\\2&5&2\\1&2&4\end{bmatrix} \;\;\;
c = \begin{bmatrix}-8\\-3\\-3\end{bmatrix}$$
We perform a similar process for matrix $A$ and vector $b$
$$Ax = b \Leftrightarrow
\begin{bmatrix}a_{11}&a_{12}&a_{13}\\a_{21}&a_{22}&a_{23}\end{bmatrix}
\begin{bmatrix}x_1\\x_2\\x_3\end{bmatrix} =
\begin{bmatrix}b_1\\b_2\end{bmatrix} \Leftrightarrow
\begin{cases}a_{11}x_1 + a_{12}x_2 + a_{13}x_3 = b_1\\a_{21}x_1 + a_{22}x_2 +
a_{23}x_3 = b_2\end{cases}$$
To make this system match the given system of equality constraints, we need to
set the components of $A$ and $b$ in the following way:
$$\begin{cases}a_{11} = 1\\a_{12} = 0\\a_{13} = 1\\a_{21} =
0\\a_{22}=1\\a_{23}=1 \\b_1 = 3 \\b_2 = 0\end{cases}$$
Therefore, we obtain the following $A$ matrix and $b$ vector:
$$A = \begin{bmatrix}1&0&1\\0&1&1\end{bmatrix} \;\;\; b = \begin{bmatrix}3\\0\end{bmatrix}$$
Then, using these $G$, $c$, $A$ and $b$ values, and using the quadratic
formulation of the problem written on the assignment
sheet, the problem is restated in the desired new form.
## Exercise 2.2
@ -118,7 +173,7 @@ The lagrangian for this problem is the following:
$$L(x, \lambda) = \frac12\langle x, Gx\rangle + \langle x, c \rangle - \lambda
(Ax - b) =$$$$= \begin{bmatrix}x_1&x_2&x_3\end{bmatrix}
\begin{bmatrix}3&0&0\\2&2.5&0\\1&2&2\end{bmatrix}
\begin{bmatrix}6&2&1\\2&5&2\\1&2&4\end{bmatrix}
\begin{bmatrix}x_1\\x_2\\x_3\end{bmatrix} +
\begin{bmatrix}x_1&x_2&x_3\end{bmatrix}
\begin{bmatrix}-8\\-3\\-3\end{bmatrix} - \lambda
@ -132,9 +187,11 @@ The KKT conditions are the following:
First we have the condition on the partial derivatives of the Lagrangian w.r.t.
$X$:
$$\nabla_x L(x, \lambda) = Gx + c - A^T \lambda = \begin{bmatrix}3 x_1 - 8 +
\lambda_1\\ 2x_1 + 2.5 x_2 - 3 + \lambda_2\\x_1 + 2x_2 + 2x_3 - 3 + \lambda_1
+ \lambda_2\end{bmatrix} > 0$$
$$\nabla_x L(x, \lambda) = Gx + c - A^T \lambda =
\begin{bmatrix}
6 x_1 + 2 x_2 + x_3 - 8 + \lambda_1\\
2 x_1 + 5 x_2 + 2 x_3 - 3 + \lambda_2\\
1 x_1 + 2 x_2 + 4 x_3 - 3 + \lambda_1 + \lambda_2\end{bmatrix} = 0$$
Then we have the conditions on the equality constraint:
@ -168,18 +225,24 @@ The KKT conditions are the following:
$$x \geq 0$$
4. The lagrangian multipliers for inequality constraints are non-negative:
$$s \geq 0$$
5. The complementarity condition holds (here considering only inequality constraints,
since the condition trivially holds for equality ones):
$$s^T x \geq 0$$
5. The complementarity condition holds (here considering only inequality
constraints, since the condition trivially holds for equality ones): $$s^T x
\geq 0$$
## Exercise 3.2
We define the dual problem is the following way:
$$\max b^T \lambda \;\; \text{ s.t. } \;\; A^T \lambda \leq c \;$$
We then introduce a slack variable $s$ to find the equality and inequality
constraints:
$$\max b^T \lambda \;\; \text{ s.t. } \;\; A^T \lambda + s = c \; \text{ and
}\; s \geq 0$$
To convert this maximization problem in a minimization one, we flip the sign of
To convert this maximization problem in a minimization one (in order to achieve
standard form), we flip the sign of
the objective and we find:
$$\min - b^T \lambda \;\; \text{ s.t. } \;\; A^T \lambda + s = c \; \text{ and
@ -187,22 +250,21 @@ $$\min - b^T \lambda \;\; \text{ s.t. } \;\; A^T \lambda + s = c \; \text{ and
We then compute the Lagrangian of the dual problem:
$$L(\lambda, x) = -b^T \lambda + x^T (A^T \lambda + s - c) - x^T s = - b^T \lambda + x^T (A^T \lambda - c)$$
$$L(\lambda, x) = -b^T \lambda + x^T (A^T \lambda + s - c) - x^T s = - b^T
\lambda + x^T (A^T \lambda - c)$$
The KKT conditions are the following:
1. The partial derivative of the lagrangian w.r.t. $x$ is 0:
$$\nabla_{\lambda} L(\lambda, x) = - b^T + x^T A^T = 0 \Leftrightarrow Ax = b$$
2. Equality constraints hold:
$$A^T \lambda + s = c$$
3. Inequality constraints hold:
$$c - A^T \lambda \geq 0 \Leftrightarrow s \geq 0 \;\;\; \text{ using 2.\ to find
that } s = c - A^T \lambda$$
4. The lagrangian multipliers for inequality constraints are non-negative:
$$x \geq 0$$
5. The complementarity condition holds (here considering only inequality constraints,
since the condition trivially holds for equality ones):
$$x^T s \geq 0 \Leftrightarrow s^T x \geq 0$$
1. The partial derivative of the lagrangian w.r.t. $x$ is 0: $$\nabla_{\lambda}
L(\lambda, x) = - b^T + x^T A^T = 0 \Leftrightarrow Ax = b$$
2. Equality constraints hold: $$A^T \lambda + s = c$$
3. Inequality constraints hold: $$c - A^T \lambda \geq 0 \Leftrightarrow s \geq
0 \;\;\; \text{ using 2.\ to find that } s = c - A^T \lambda$$
4. The lagrangian multipliers for inequality constraints are non-negative: $$x
\geq 0$$
5. The complementarity condition holds (here considering only inequality
constraints, since the condition trivially holds for equality ones): $$x^T s
\geq 0 \Leftrightarrow s^T x \geq 0$$
Then, if we compare the KKT conditions of the primal problem with the ones above
we can match them to see that they are identical: