Revisiting the Approximate Carath'eodory Problem via the Frank-Wolfe Algorithm

The approximate Carath'eodory theorem states that given a compact convex set Csubsetℝn and p∈left[2,+∞right[, each point x^*∈C can be approximated to ε-accuracy in the ℓp-norm as the convex combination of O(pDp2/ε2) vertices of C, where Dp is the diameter of C in the ℓp-norm. A solution satisfying these properties can be built using probabilistic arguments or by applying mirror descent to the dual problem. We revisit the approximate Carath'eodory problem by solving the primal problem via the Frank-Wolfe algorithm, providing a simplified analysis and leading to an efficient practical method. Furthermore, improved cardinality bounds are derived naturally using existing convergence rates of the Frank-Wolfe algorithm in different scenarios, when x^* is in the interior of C, when x^* is the convex combination of a subset of vertices with small diameter, or when C is uniformly convex. We also propose cardinality bounds when p∈left[1,2right[cup+∞ via a nonsmooth variant of the algorithm. Lastly, we address the problem of finding sparse approximate projections onto C in the ℓp-norm, p∈left[1,+∞right].