S710: Statistical Computing
S710 provides graduate students in statistical science and related disciplines with a solid introduction to three areas of computational mathematics: numerical linear algebra, numerical optimization, and stochastic simulation. Although the choice of applications (and therefore, to some extent, the choice of topics) is guided by the concerns of statistics, this course is suitable for any student desiring a survey of these areas. Please note that S710 is NOT a course on computer packages for data analysis.
Class time: Monday, Wednesday : 9:30-10:45
Capacity: As many as needed.
Instructor: Daniel Manrique-Vallier - firstname.lastname@example.org
Other Contact(s): Michael Trosset October 17, 2012
Sequence: S710 is not part of a sequence.
Prerequisites: See Matrix Algebra, Calculus, and Computing.
Algebra Required?: Students should have some familiarity with basic concepts and properties of vector spaces and matrices, e.g., linear independence of vectors, eigenvalues of matrices. Matrix notation is used extensively. Students are expected to understand proofs presente
Calculus Required?: Students should have some familiarity with basic concepts from vector calculus, e.g., directional derivatives. Students are expected to understand derivations presented in class and/or reading assignments. Some homework problems may involve derivations.
Day(s) per week offered: It has been taught on MWF in 50-minute class meetings. It could also be taught on TR in 75-minute meetings. There is no computer lab.
Recommended follow-up classes: S710 provides background in computational mathematics that will benefit students in a number of courses, e.g., STAT-S 631 and STAT-S 675.
Substantive Orientation: Statistics, informatics, computer science, library & information science, mathematics, physics, chemistry.
Statistical Orientation: Not applicable.
Books used: Numerical Linear Algebra, by L.N. Trefethen and D. Bau. Numerical Optimization, by J. Nocedal and S.J. Wright. Introducing Monte Carlo Methods with R, by C.P. Robert and G. Casella.
Applied/Theoretical: The major topics themselves (numerical linear algebra, numerical optimization, stochastic simulation) will be of interest to graduate students in statistics, informatics, and computer science. Students interested in practical methods for data analysis wil
Software Used: R
How the software is used: Some problems require programming, typically implementing and/or using an algorithm studied in class.
Problem Sets: Weekly.
Data Analysis: Some problems, e.g., solving a nonlinear least squares problem by Gauss-Newton, involve data; however, S-710 is not a course about data analysis.
Presentations: Yes, typically a paper on a topic related to the topics studied in class.
Keywords: computational mathematics, numerical linear algebra, numerical optimization, stochastic simulation