MDS reveals the structure of the data by providing a visual presentation of similarities among a set of objects. In more detail, MDS places an object in an n-dimensional space, where the distances between pairs of points corresponds to the similarities among the pairs of objects. Usually, the dimensional space is a two-dimensional Euclidean space, but it may be non-Euclidean and have more than two dimensions. In accordance with the meaning of the input matrix, MDS can be mainly categorized into two types: metric MDS, where the input matrix is metric-based, and nonmetric MDS, where the input matrix is nonmetric-based.
Metric MDS is also known as principal coordinate analysis, which first transforms a distance into similarities. ...
