Math Problem Statement
Which of the following is/are true for PCA and MDA?
L Projection direction may not be the best in terms of minimum squared error for MDA
II. Projection direction is best in terms of minimum squared error for PCA.
a) Only I
b) Only II
c) Both I and II
d) Neither I nor II
Oa)
b)
Oc)
Od)
1 point
For a two-class problem. if \mathcal{S}{W} is within class scatter. \mathcal{S}{B} is between class scatter and mu_{1} and mu_{2} are the corresponding means. Which of the following is correct for direction of projection using MDA?
a) S_{B} ^ - 1 * (mu_{1} - mu_{2})
b) S_{w}(mu_{1} - mu_{2})
c) S_{w} ^ - 1 * (mu_{1} - mu_{2})
d) None of the above
Solution
Ask a new question for Free
By Image
Drop file here or Click Here to upload
Math Problem Analysis
Mathematical Concepts
Principal Component Analysis (PCA)
Multiple Discriminant Analysis (MDA)
Linear Algebra
Formulas
PCA: Principal components maximize variance
MDA: Optimal projection direction: w* = S_W^-1 * (μ_1 - μ_2)
Theorems
Eigenvalue Decomposition
Fisher's Linear Discriminant
Suitable Grade Level
Undergraduate
Related Recommendation
Understanding True Statements About PCA and MDA: Minimizing Squared Error
PCA Projection Direction for Data Points on Elliptical Curves
Understanding PCA Projection Directions with Elliptical Data
Information Loss in PCA with Roughly Equal Eigenvalues & True Statements about PCA
Understanding PCA Projections: Direction of Maximum Variance in Ellipse Data