|
Examples of SMACK research projects |
|
Mathematical theory of classification: We develop principles for probilistic predictive unsupervised, semi-supervised and supervised classification, as well as stochastic computation algorithms for practical applications. Modelling of genetic population structure: We develop statistical methods (models and estimation algorithms) for inferring genetic population structure and identification of migration. These methods have been applied to a wide variety of species within different contexts, from genetic epidemiology to metapopulation ecology. Several subprojects exist in collaboration with biologists. Models and methods in bioinformatics: We work within many fields of bioinformatics, from sequence analysis to studies of protein structure. Graphical models: Graphical models are used e.g. in artificial intelligence to describe complex systems. Our research is connected to both further development of graphical models and application of these to real systems. Stochastic computation methods: Modern statistics utilizes stochastic computation as a central tool for model fitting and structural learning. We develop novel algorithms mainly within the class of so called Markov chain Monte Carlo -methods. |
