Sequential Monte Carlo methods (particle filtering/smoothing) for a toy problem of the following form
- x(t+1) = a x(t) + v(t), v(t) ~ N(0,sigmav^2),
- y(t) = c x(t) + e(t), e(t) ~ N(0,sigmae^2)
where a and c denote scale parameters and the noise variances are given by sigmav^2 and sigmae^2.
The following files are included
- toyex_pf: estimates the states given a data realisation and the parameters using a bootstrap particle filter.
- toyex_bpf: estimates the states given a data realisation and the parameters using a information filter (backward particle filter).
- toyex_fl: estimates the states given a data realisation and the parameters using a fixed-lag smoother.
- toyex_ffbsm: estimates the states given a data realisation and the parameters using a forward-filtering backward-smoothing.