The Hidden Markov Model is a standard model of statistics whose domains of applications span a variety of topics, from data compression and speech recognition to bioinformatics. Perspectives on Hidden Markov Models is a course given as a collaboration between the departments of mathematics, theoretical statistics, and computer science and the centre for bioinformatics (BirC). The purpose of the course is to let researchers from the different departments present their (pure and/or applied) perspective on the topic.
The course is given as a series of six lectures throughout the fourth quarter of 2005/2006, i.e., the second half of the spring semester of 2006. It is a course of the graduate school of the Center for theory in Natural Science and is open to all graduate students at the faculty of science. The prerequisite for taking the course is a certain amount of mathematical maturity and familiarity with basics of mathematical modeling (as might be achieved through a bachelor degree at the Faculty of Science).
The course is not graded. For passing the course (worth 5 ECTS), an assignment has to be completed in a satisfactory way.
To sign up for the course, please email Peter Bro Miltersen as soon as possible and preferably before the start of the course.
Related courses:April 7: Christian Storm Nørgaard Pedersen, Computer Science Department and BirC. Comparison of Hidden Markov Models. Reading: Paper 1 og Paper 2. Slides. Project suggestions.
April 21: Søren Asmussen, Department of Theoretical Statistics. Markov-modulation in applied probability. Notes.
April 28: Peter Bro Miltersen, Computer Science Department. Hidden Markov models in data compression and the complexity of global maximum likelihood computation. Second part of the lecture will be based in part on this paper.
May 5: Niels Lauritzen, Department of Mathematical Sciences. Parametric analysis of hidden Markov Models. Abstract. Paper. Paper. Project Suggestion.
May 19: Jens Ledet Jensen, Department of Theoretical Statistics. Asymptotics for Hidden Markov Models. Abstract.
May 26: Søren Asmussen, Department of Theoretical Statistics. Markov chain Monte Carlo and HMMs. Ole F. Christensen, BirC. Modelling conflicting primate phylogenies in DNA sequences using a phylogenetic hidden Markov model.