Feasibility and a fast algorithm for Euclidean distance matrix optimization with ordinal constraints (Q2181603)

The authors study the Euclidean distance matrix optimization with ordinal constraints (EDMOC) and propose the feasibility of EDMOC. The authors develop a majorized penalty method to solve a large-scale EDMOC model and convert a convex EDM optimization problem into an isotonic regression problem. Due to the partial spectral decomposition and the fast solver for an isotonic regression problem, the performance of the proposed algorithm is demonstrated to be superior both in terms of the solution quality and CPU-time in sensor network localization and molecular conformation.