This course is NOT running in academic year 201415
Please see CY902N
 To familiarise students with the techniques of Computational Linear Algebra.
 Understand the performance issue associated with Computational Linear Algebra.
 Familiarise students with standard Linear Algebra Libraries.
 Learn to formulate problems in terms of Linear Algebra operations.
 To introduce the central mathematical ideas behind algorithms for the numerical solution of Optimization problems.
 To provide theoretical justification for various algorithms and to outline basic issues of numerical analysis involved in Optimization problems.
 To present key methods for both constrained and unconstrained Optimization.
 To learn about available implementations and software packages.
 To learn about various heuristic strategies used in science and engineering applications, in particular for global Optimization.

Learning Outcomes:
 Use standard techniques of Computational Linear Algebra.
 Use standard software for Computational Linear Algebra.
 Formulate Scientific Computing problems as Linear Algebra Operations.
 Understand the Mathematical background to solving Numerical Optimization problems.
 Understand and use methods for constrained and unconstrained Optimization.
 Demonstrate familiarity with standard software package for Optimization.

Prerequisites:
 A good working knowledge of a scientific programming language (as for example taught in PX270 C Programming).
 A background in Linear Algebra (as taught in MA251).
 Some knowledge of Vector Calculus (as taught in MA231).
Syllabus:
 Linear Algebra in Computational Science
 Computer Architecture and Linear Algebra
 Revision of Linear Algebra
 Basic Linear Algebra routines
 Small matrix eigenvalue problem and linear equations
 Large matrix eigenvalue problems and linear equations
 Singular Value Decomposition
 Programming techniques in Linear Algebra
 Mathematical Introduction to Optimization
 Optimality Conditions
 Unconstrained Optimization Algorithms
 Computational Issues
 Global Optimization
 Constrained Optimization Algorithms
 Optimal Control and Dynamic Programming
 More Computational Issues
