Implement missing dimensionality reduction algorithms

by Sergey Lisitsyn for Shogun Machine Learning Toolbox (Technical University Berlin / Max Planck Campus Tübingen)

Dimensionality reduction is the process of finding a suitable low-dimensional dataset from high-dimensional one by reducing its dimensionality. One of the most important practical issues of applied machine learning, it is widely used for preprocessing real data. Its importance is deriving from challenges presented by high-dimensional spaces, performance and other issues. Being large-scale, SHOGUN is missing preprocessors for dim. reduction and this project aims for implementation some of them.