From: Sequential Monte Carlo for inference of latent ARMA time-series with innovations correlated in time
PF type | State estimation error (MSE) | |
---|---|---|
Known \(\sigma _{u}^{2}\) | Unknown \(\sigma _{u}^{2}\) | |
AR(1), H=0.5 | 1.1081 | 1.1945 |
AR(1), H=0.7 | 1.3946 | 1.4397 |
AR(1), H=0.9 | 1.1195 | 1.1970 |
MA(1), H=0.5 | 1.0223 | 1.0686 |
MA(1), H=0.7 | 1.0585 | 1.1136 |
MA(1), H=0.9 | 0.87374 | 0.94053 |
ARMA(1,1), H=0.5 | 1.5947 | 1.6197 |
ARMA(1,1), H=0.7 | 1.7852 | 1.8516 |
ARMA(1,1), H=0.9 | 1.7214 | 1.7362 |