Computer Intensive Methods in Statistics, 7.5 Credits
About the course
The course covers the statistical theory of stochastic simulation (Monte Carlo methods) and other computer-intensive statistical methods, i.e. methods used to solve problems that are difficult to solve by analytical methods. The course covers the generation of random numbers from different distributions and integral estimation including error estimation. Also included are variance reducing methods, such as using antithetic variables, control variables, conditioning and stratified sampling. Emphasis is placed on methods for simulating Poisson processes to allow for simulation of queuing and inventory systems. Finally methods for constructing confidence intervals and performing hypothesis testing using non-parametric and parametric bootstrap, for both one-and two-dimensional data, are introduced. Throughout the course there is a major focus on implementing methods using appropriate software, such as Matlab, R or C.
Level of Education: Advanced
Course Menu
Contact Information
Dept of Mathematics and Mathematical Statistics
MIT-huset
901 87 Umeå
