Computational Statistics
Computational Statistics
Computational statistics concerns the development, implementation and study of computationally intensive statistical methods. Such methods are often used e.g. in the fields of data visualization, the analysis of large datasets, Monte Carlo simulation, resampling methods such as the bootstrap, permutational methods, Markov Chain Monte Carlo methods and various numerical methods of equation solving such as the EM algorithm and Newton-Raphson iteration. A very powerful tool to implement such methods is the R statistical programming language.
This course will present essential methods in computational statistics in a practical manner, using real-world datasets and statistical problems. Examples will include e.g. 1) evaluating and comparing the performance of different statistical techniques in a specific setting using simulation, 2) implementing complex methods such as an EM algorithm to fit a joint model, 3) implementing the bootstrap to obtain a standard error estimate which is not available in closed-form. We will also develop advanced R programming skills.
Course Objectives
At the end of the course, the student:
- will have developed advanced and computationally efficient R programming skills,
- is able to conduct and report on simulation studies, comparing the performance of statistical methods in specific settings,
- is able to implement and use methods for statistical inference such as the bootstrap and permutation test,
- will be familiar with the Metropolis-Hastings algorithm, as an example of a Markov Chain Monte Carlo method,
- is familiar with some widely used numerical methods,
- will be able to translate new statistical methods from the literature into a usable R program.
Target Group
Our courses are aimed at clinical researchers, nurses, general practitioners, and other health professionals who want to improve their skills in epidemiology, statistics and (clinical) research.
Computational Statistics (face to face)
1 week, fulltime