Quiz 2, ICS 273A Intro-ML1. Consider a data-cloud in the form of a pancake. We will perform a PCA on this dataand reduce the dimension from 3 to 2 dimensions. If we perform the projection€ y = U1:2Tx how does the data-cloud look like in y coordinates?A) Again a pancake (i.e. circular)B) One high variance and one low variance direction2. True or false: no two eigen-values of a symmetric, positive definite matrix can be the same? A) TrueB) False3. PCA is a A) supervised techniqueB) semi-supervised techniqueC) unsupervised technique4. The rank of a covariance matrix computed from a data-matrix X with D rows (attributes) and N columns (data-items) is: (D<N)A) DB) NC) 15. True or false: the determinant of a symmetric positive definite matrix is always positive: A) TrueB) False6. Same as 5 but now for the trace:A) True B)
View Full Document