CS 451 Quiz 20
Principal Component Analysis (PCA)
What are typical applications of PCA?
Check all that apply.
PCA allows us to discard redundant features
PCA projects p-dimensional data onto a q-dimensional subspace, where q < p. What is p?
the number of data points
the number of features
the number of centroids
PCA computes the same thing as
PCA finds a lower-dimensional subspace minimizing
the maximum distance between a data point and its projection
the variance of the projections of the data points
the sum of the squared distances between the data points and their projections
In linear regression, one of the coordinate axes is "special", while PCA treats all axes equally
What preprocessing is necessary before running PCA?
Both mean normalization and feature scaling are necessary
Mean normalization is necessary; feature scaling is optional
Mean normalization is optional; feature scaling is necessary
Both mean normalization and feature scaling are optional
The PCA algorithm computes the eigenvectors of Sigma, where Sigma is the
PCA can be computed via the Matlab command [U, S, V] = svd(Sigma). The resulting k-dimensional subspace is spanned by
the pseudo-inverse of S
the first k columns of U
the first k rows of V
What does SVD stand for?
Software Version Descriptor
Structural Valve Deterioration
Singular Vector Differentiation
Special Value Division
Single Vision Distance
Strategic Vertical Differentiation
Singular Value Decomposition
Signature Verification Data
Strategic Vision Development
Feedback on working in teams
(not part of your quiz grade)
When working on homework problems, which do you prefer?
Working in teams
If teams are mandatory on a homework problem, which do you prefer?
Choosing team members myself
Teams assigned by the professor
Either is fine
This content is neither created nor endorsed by Google.
Terms of Service