5630 Numerical Optimization
Spring 2009
Instructor: 	Allan Struthers
Office: 		Fisher 212
Phone: 		487-3541
e-mail: 		struther@mtu.edu
briefcase:	MA5630 on Zimbra
Topics
 Numerical solution of unconstrained and constrained optimization problems and nonlinear equations.
Topics include: 
Optimality conditions;
Local convergence of Newton and Quasi - Newton methods;
Line search and trust region globalization techniques;
Quadratic penalty and augmented Lagrangian methods for equality - constrained problems;
Logarithmic barrier method for inequality-constrained problems; and 
Sequential Quadratic Programming.
Text
Numerical Optimization 2 nd edition 
Jorge Nocedal and Stephen Wright.
ISBN : 0387303030
Coding 
You will be expected to implement many of the algorithims discused in the class. You can use MathLab, Mathematica, C, C++, Java, or Fortran. I will be able to provide the most help with Mathematica and C. There will be some students in the class with little programming experience. I will run a "Programming Boot Camp" for these students at the start of the course  
Prerequisites : MA4330 or MA4610 or MA5627 or MA4630 or consent of instructor.
Meeting Times
The course meets M/W/F 11 - noon.