The goal of multidimensional scaling (MDS) is, given pairwise dissimilarities (i.e. a matrix of distances $D = (d_{ij})$), find coordinates $x_i, x_j$ such that:
$$d_{ij} \approx ||x_i - x_j||_2$$