We work on novel sampling techniques to design faster algorithms for solving problems on large data. Algorithmic applications of our methods include dimensionality reduction, clustering, and linear algebraic problems on large matrices.