3450:635 Project #1, Optimization, Dr. Norfolk
Due:
Great Circle Distance
Using the input le of the locations of the major airports in each State, (latitudes
are North, longitudes are West), write a routine to compute the Great Circle distance
between any
Midterm Optimization 3450:635
Dr. Norfolk Due : Mar 26 , 2008
Show your work.
Quote any references.
You may only discuss this test with me.
Each hint costs 1 point.
1. Suppose that we wish to nd a value x that simultaneously solves f (x) = 0, g(x) = 0
Final Examination
Optimization 3450:635 Dr. Norfolk
Due : May 5 , 2008
Show your work.
Quote any references.
You may only discuss this test with me.
Each hint costs 1 point.
For questions (13), dene
1
f (x) = [39, 12]x + xT
2
3 1
1
2
x + c1 x4 + c2 x4
Quasi-Newton Methods
Quasi-Newton methods are a class of optimization methods which mimic Newtons method
without computing second derivatives.
The basic idea is to approximate the Hessian B 2 f (xk ) or its inverse H (
using a rank 1 or rank 2 update at e
Final Examination
Optimization 3450:635
Dr. Norfolk
Due : 12:00pm, May 3 , 2010
Show your work.
Quote any references.
You may only discuss this test with me.
Each hint costs 1 point.
1. Fertilizer comes in 50-pound bags. Each bag is labelled with a tr
3450:635 Project #2, Optimization, Dr. Norfolk
Due:
Travelling Salesperson Tour
Using the Great Circle calculations from project #1, and an algorithm of your
choice, nd the best tour of all cities that you can, starting and ending in Washington,
DC.
Your
3450:635 Project #4, Optimization, Dr. Norfolk
Due:
Newtons Method
Implement the Newtons Method with Choleskis method and Hessian modication, step length = 1.0, and objective function given by Rosenbrocks Banana:
f (x1 , x2 ) = 100(x2 x2 )2 + (1 x1 )2
1
T
3450:635 Project #3, Optimization, Dr. Norfolk
Due:
Steepest Descent
Implement the steepest descent algorithm, with back-tracking line search starting
at = 1.0, and objective function given by Rosenbrocks Banana:
f (x1 , x2 ) = 100(x2 x2 )2 + (1 x1 )2
1
T
Midterm
Optimization 3450:635
Dr. Norfolk
Due : Mar 26 , 2010
Show your work.
Quote any references.
You may only discuss this test with me.
Each hint costs 1 point.
1. A class is given the problem of nding a point on the intersection of the surfaces f
Convex Optimization
(EE227A: UC Berkeley)
Lecture 14
(Gradient methods II)
07 March, 2013
Suvrit Sra
Organizational
Take home midterm: will be released on 18th March 2013
on bSpace by 5pm; Solutions (typeset) due in class, 21st
March, 2013 no exceptions!