Dimensionality Reduction - Mining of Massive Datasets - page 414

[7] J. Sun, Y. Xie, H. Zhang, and C. Faloutsos, “Less is more: compact matrix decomposition for large sparse graphs,”
Proc. SIAM Intl. Conf. on Data Mining , 2007.
[8] M.E. Wall, A. Reichtsteiner and L.M. Rocha, “Singular value decomposition and principal component analysis,”
in A Practical Approach to Microarray Data Analysis (D.P. Berrar, W. Dubitzky, and M. Granzow, eds.), pp. 91-
109, Kluwer, Norwell, MA, 2003.
1 Recall M i denotes multiplying by the matrix M i times, as discussed in Section 5.1.2 .
2 Note that a stochastic matrix is not generally symmetric. Symmetric matrices and stochastic matrices are two classes
of matrices for which eigenpairs exist and can be exploited. In this chapter, we focus on techniques for symmetric
matrices.
3 Note that Fig. 11.7 shows V T , while this multiplication requires V .
4 In Fig. 11.7 , it happens that U and V have a significant number of 0's. However, that is an artifact of the very regular
nature of our example matrix M and is not the case in general.