Sequential Monte Carlo methods
Schedule & Material
Most of the scheduled time will consist of traditional lectures. Slides will be provided via this page, but note that the blackboard will be used quite extensively as well.
Each day, a few hours of practical sessions are also scheduled. Assistants will be available to help you with the exercises and hand-in assignments during these sessions. A large share of the problems will be on implementation, so please bring your own laptop with a suitable language of your choice installed (e.g., Python, Julia, R, Matlab, ...). (Note! You will probably need to spend more time than the scheduled sessions to complete the assignments.)
|9.15-12.00||Hambergsalen, Geocentrum||Lecture 1-3||1. Introduction and probabilistic modelling; 2. Probabilistic modelling of dynamical systems and the filtering problem; 3. Monte Carlo and importance sampling||Le1, Le2, Le3|
|13.15-15.00||Hambergsalen, Geocentrum||Lecture 4-5||4. The bootstrap particle filter; 5. Convergence of bootstrap PF||Le4, Le5|
|15.15-17.00||Norrland I & II, Geocentrum||Practicals||Preparatory problems, Exercises set I|
|9.15-12.00||Hambergsalen, Geocentrum||Lecture 6-8||6. Auxiliary variables and the auxiliary PF; 7. the fully adapted PF; 8. Path space view, path degeneracy and ESS||Le6, Le7, Le8|
|13.15-15.00||Hambergsalen, Geocentrum||Lecture 9-10||9. Parameter learning and likelihood estimation; 10. The particle filter as a likelihood estimator||Le9, Le10|
|15.15-17.00||Norrland I & II, Geocentrum||Practicals||Exercises set II|
Social activity. More information later.
|9.15-12.00||Room I and XI, University main building||Practicals||Exercises set III|
|13.15-17.00||Room IV, University main building||Lecture 11-14||11. Metropolis-Hastings; 12. Particle Metropolis-Hastings; 13. Gibbs sampling; 14. Particle Gibbs||Le11, Le12, Le13, Le14, MwG example|
|9.15-11.00||Room I and XI, University main building||Practicals||Exercises set IV|
|11.15-13.00||Room IV, University main building||Lecture 15-16||15. General SMC; 16. SMC samplers||Le15, Le16, SMC sampler notes|
|15.15-17.00||Room IV, University main building||Lecture 17-18||17. SMC for probabilistic programming; 18. High-dimensional SMC||Le17, Le18|