Sequential Monte Carlo Projects

The starred projects are (in my opinion) a bit harder than the rest.

Methodology
  1. Sensitivity methods for SMC with application to recursive parameter estimation
  2. Recursive parameter estimation using on-line Expectation-Maximization methods
  3. Performance analysis of recursive Bayesian parameter estimation
  4. Adaptive SMC samplers for global optimization
  5. Adaptive SMC samplers for Bayesian variable selection
  6. Trans-dimensional SMC samplers for partially observed point processes
  7. Comparison of smoothing algorithms for state-space models
  8. * SMC methods for optimal control 

Theory
  1. Uniform convergence bounds
  2. Central limit theorems
  3. Exact calculations of asymptotic variance for SMC methods for a special class of dynamic models
  4. * Convergence results for non-bounded test functions
  5. Central limit theorems for smoothing algorithms
 Applications
  1. SMC methods for infinite hidden Markov models (nonparametric Bayes)
  2. SMC methods for changepoint models
  3. SMC methods for hierarchical clustering via coalescent