A proximal Newton framework for composite minimization: Graph learning without Cholesky decompositions and matrix inversions.

Quoc Tran-Dinh, Anastasios Kyrillidis, Volkan Cevher

Browse the full ICML paper archive.