Mini Course 2 - Eric Moulines (CNRS/ENST)

Inference for Stochastic Processes

1. Hidden Markov Models: from linear Gaussian state-spaces to non-linear models
2. The Kalman filter
3. An introduction to the particle filter
4. The Baum-Welch algorithm
5. Particle filter: some convergence results
6. The auxiliary filter
7. Parameter estimation I: theory and methods
8. Parameter estimation II: algorithms
9. The Filtering problem: stability, ergodicity

Click here to download the notes.
Click here to download the exercises.
To be also discussed: Limit theorems for weighted samples with applications to sequential monte carlo methods.