Dimensionality ¡Reduc1on ¡ Lecture ¡23 ¡
David ¡Sontag ¡ New ¡York ¡University ¡
Slides adapted from Carlos Guestrin and Luke Zettlemoyer
Dimensionality Reduc1on Lecture 23 David Sontag New York - - PowerPoint PPT Presentation
Dimensionality Reduc1on Lecture 23 David Sontag New York University Slides adapted from Carlos Guestrin and Luke Zettlemoyer Assignments Last homework assignment released tonight,
Slides adapted from Carlos Guestrin and Luke Zettlemoyer
2
Slide from Yi Zhang
i
i
i
i
From notes by Andrew Ng
1 m
m
(x(i)Tu)2 = 1 m
m
uTx(i)x(i)Tu = uT
m
m
x(i)x(i)T
Let x(i) be the ith data point minus the mean. Choose unit-length u to maximize: Let ||u||=1 and maximize. Using the method of Lagrange multipliers, can show that the solution is given by the principal eigenvector of the covariance matrix! (shown on board)
T ¡Xc ¡
23
In high-dimensional problem, data usually lies near a linear subspace, as noise introduces small variability Only keep data projections onto principal components with large eigenvalues Can ignore the components of lesser significance. You might lose some information, but if the eigenvalues much
5 10 15 20 25 PC1 PC2 PC3 PC4 PC5 PC6 PC7 PC8 PC9 PC10 Variance (%)
Slide from Aarti Singh
Principal components: