Computational Statistics

3 Credits · Level: Undergraduate

Prerequisites: MATH 2140 or ECON 2110 . Description: Modern statistics increasingly relies on computationally intensive methods and techniques. This course introduces students to some of those commonly used methods. It assumes knowledge of concepts from introductory mathematical statistics including probability distributions, expectations, maximum likelihood estimation, and hypothesis testing. While some familiarity with the open source software R is preferred, no prior programming experience is expected. Topics will include random number generation, deterministic methods like Fisher scoring, rejection sampling, importance sampling, permutation tests, jackknifing, cross validation, bootstrap resampling, elementary Bayesian statistics, Gibbs sampling, as well as the expectation-maximization (EM) algorithm and variational inference. All computation will be performed in R.