Something PCA is really good at seems to be defining some sort of "dimension" in very abstract data. For example, political opinions. I saw something on Twitter where a data scientist had made a PCA that let you compare your policy opinions to those of economists and told you who you were most similar to.
Something PCA is really good at seems to be defining some sort of "dimension" in very abstract data. For example, political opinions. I saw something on Twitter where a data scientist had made a PCA that let you compare your policy opinions to those of economists and told you who you were most similar to.