Saved Question 9 (1 point) In the Principal Component Analysis (select all that are true): The principal directions are the directions in the features space along which the data vary the most. The principal components provide the low-dimensional linear surfaces that are farthest to the observation. The first principal component optimizes max 911...910 G ( 91j j=1 i=1 where, x1 = (x11 ...)? represents the vector of observation 1, similarly 91 = (911 ... 91D)? represents the first principal vector and the column means of NxD data set X are zero