Mathematical functions for dimension reduction in big data

284 Views Asked by At

I need a Mathematical function that can project a high dimensional data set into a low dimensional one.

in fact this topic is very important to me because of my Master thesis's proposal. All approaches likes PCA,Gaussian random projection and others have been used before, and i have got confused to what to do!

what kind of functions do you suggest? or what kind of combination of existing methods do you suggest?