TMA4300 Computer Intensive Statistical Methods

The curriculum is covered in the lectures and the exercises. As reference text we will use:

Givens, G. H. and Hoeting, J. A. "Computational Statistics", Second Edition, 2013, John Wiley & Sons, New Jersey.

You can download this book electronically, i.e. as an e-book, from the NTNU library, when you are in the NTNU network. This book captures most of the topics discussed in the lecture and gives a good introdution into computational statistics. Additional references to specific topics might be used throughout the semester.

The lecture will consist of three parts:

1. Algorithms for stochastic simulation
2. Markov chain Monte Carlo Methods
3. Expectation-maximization algorithms, bootstrap and classfication methods

that are interchanging with exercise periods.

For part 1 you may also want to have a look at chapters 1 and 2.1-2.4 of the book

"Markov chain Monte Carlo: Stochastic simulation for Bayesian inference" by Gamerman and Lopes, 2nd edition Chapman & Hall/CRC 2006.

which give a bit more details in particular regarding simulation from discrete and continuous quantities. Further relevant sections will be noted on the lecture plan

In part 3 we will talk about classification. However, this is not covered in neither of the books before. We will therefore use

"The elements of statistical learning" by Hastie, Tibshirani and Friedman, 2nd edition.

The relevant section are noted on the lecture plan