In this problem you will examine a method for nonlinear dimensionality reduction known as locally linear embedding (LLE). Let x1, . . . , xn Rq. The idea behind LLE is that even though the data points may lie on a nonlinear structure/manifold, at least locally the data appear linear. Therefore, it should be possible to represent each data point as