In this presentation, we obtain some new convergence results for the Simplicial Method.

To this end, we use the duality between the Generalized Conditional Gradient and the Simplicial Method. These two algorithms are popular convex optimization algorithms, and they are dual to each other in the sense that they produce the same candidates.

These two popular convex optimization algorithms will be presented at the beginning of the talk.

 

We will conclude this presentation with a numerical example on a variant of the Support Vector Machine problem, which is a linear classification problem.