In this paper, a review of cone programming formulation of soft margin semi-supervised support vector machines is provided. The S 3 VM is known to be NP-hard, thus their cone programming reformulation remains NP-hard. However, the reformulation converts the problem into a convex optimization problem. The formulations can be classified into semidefinite programming reformulation and copositive reformulation. We have collected several semi-definite and copositive programming reformulations. The relations between these reformulations are also discussed.
|Title of host publication||2019 2nd International Conference on Computing, Mathematics and Engineering Technologies, iCoMET 2019|
|Number of pages||6|
|Publication status||Published - 2019|