1. Fisher’s linear discriminant is
wˆ T Sbwˆ
wˆ = arg max J(wˆ ) = arg max wˆ T S wˆ ,
wˆ wˆ w
where Sb = (m1 − m2)(m1 − m2)T and Sw = Σj Σα(xα − mj)(xα − mj)T .
(a) By writing ∂J
= 0, show that
S−w1Sbwˆ = J(wˆ )wˆ .
(b) Explain why wˆ ∗ is the eigenvector for which J(wˆ ) is the maximum eigenvalue of S−w1Sb.
(c) Explain why Sbwˆ
is always in the direction of m1 − m2 and thus show that
wˆ ∗ = const. • S−w1 (m1 − m2).
2. Another way to optimize Fisher’s linear discriminant (suggested by Barry Fridling):
(a) Show that for any two real vectors x and y
(xT y)2 ≤ (xT x)(yT y), (Cauchy-Schwarz).
(b) Show that if the λk are positive,
Using the Optdigits dataset from the UCI repository, implement PCA. Reconstruct the digit images and calculate the reconstruction error E(n) = j ||xˆj − x||2 for various values of n, the number of eigenvectors. Plot E(n) versus n.
DescriptionIn this final assignment, the students will demonstrate their ability to apply two ma
Path finding involves finding a path from A to B. Typically we want the path to have certain properties,such as being the shortest or to avoid going t
Develop a program to emulate a purchase transaction at a retail store. Thisprogram will have two classes, a LineItem class and a Transaction class. Th
1 Project 1 Introduction - the SeaPort Project series For this set of projects for the course, we wish to simulate some of the aspects of a number of
1 Project 2 Introduction - the SeaPort Project series For this set of projects for the course, we wish to simulate some of the aspects of a number of