SUMT - Introduction | Previous | Next |
To illustrate SUMT, consider the example for the Frank-Wolfe algorithm presented in the textbook (Section 13.9) | |
![]() |
![]() |
The feasible region is shown to the right (and in Fig 13.17). As the text indicates, the optimal solution turns out to be (1, 1.5), whereas the unconstrained maximum is (2.5, 2). To apply SUMT, we form the barrier function | |
![]() |
|
and the corresponding unconstrained optimization problem, | |
![]() |