MATH30003 Approximation Theory and Numerical Integration SEMESTER: First
CONTACT: Dr Ruth Thomas (M/O16) CREDIT RATING: 10
Aims:
  • To introduce some methods for approximating a function by polynomials, splines and rational functions.
  • To introduce the concept of best approximation.
  • To introduce Gaussian quadrature rules.
Intended Learning Outcomes: On successful completion of this module students will:
  • Have acquired active knowledge and understanding of basic approximation theory (including the formulation of some best approximation problems) and of an advanced technique in numerical integration.
  • Be able to approximate a function by a polynomial, piecewise polynomial or Padé approximant.
Pre-requisites: 153, 157, 211 (ex-UMIST), MT1202 Sequences and Series (ex-VUM)

This course is NOT available to students who have taken MT2181 in previous years.

Dependent Courses:  
Course Description: The first half of the course is concerned with approximation theory. The aim is to approximate a complicated function by a much simpler function (such as a polynomial), which is easier to evaluate, differentiate and integrate. In numerical integration, the course builds on the ideas introduced in Module 157. A family of integration rules, known as Gaussian quadrature rules, is introduced.
Teaching Mode: 2 Lectures per week
1 Tutorial per week
Private Study: 5 hours per week
Recommended Texts: E, Suli and D Mayers.  An Introduction to Numerical Analysis.  CUP, 2003.
Assessment Methods: Coursework: 20%
Coursework Mode: Test in Week 7: Students will be asked to write out their solutions to a small number of problems chosen from a set notified in advance.
Examination: 80%
Examination is of 2 hours duration at the end of the First Semester.
No of lectures: Syllabus
2 Introduction and revision of function norms; Weierstrass' Theorem for polynomial approximation.
2 Best minimax (L \infty) polynomial approximation: the equioscillation property and de la Vallée Poussin's Theorem.
3 Chebyshev polynomials: definition and properties. Best L \infty approximation to x n+1 by a polynomial of degree n or less. Economisation of power series. Lanczos \tau method.
4 Best Euclidean (L2) polynomial approximation: construction and properties of orthogonal polynomials (the Gram-Schmidt process).
2 Introduction to Hermite interpolation.
3 Approximation by piecewise polynomials (splines): definition and construction.
2 Rational approximation: Padé approximants.
4 Gaussian quadrature.

 Last revised August, 2006