what is the difference between PCA and MDS analysis
In: 0
PCA is a rotation transformation. It twists the data in whatever number of dimensions it has so that the direction where the data has the most variance is the first dimension. It’s like turning an American football in your hand so that the long axis is pointing a certain way. It also turns the second most varying direction to be doing the second axis. The third most varying to the third, etc.
MDS tries to position points in a lower dimension space to reflect how far apart they are in higher dimensions. It’s like you have a shape made of springs connected together and you press it flat on a table.
Latest Answers