### Table 1: MC lters: linear Gaussian model

2000

"... In PAGE 25: ... Results For the Kalman lter, we obtain pV AR xkjk = 0:79. For the di erent MC lters, the results are presented in Table1 and Table 2. With N = 500 trajectories, the estimates obtained using MC methods are similar to those obtained by Kalman.... ..."

Cited by 293

### Table 1. MC filters: linear Gaussian model

2000

Cited by 293

### Table 1: MC filters: linear Gaussian model

2000

"... In PAGE 25: ...79. For the different MC filters, the results are presented in Table1 and Table 2. With N = 500 trajectories, the estimates obtained using MC methods are similar to those obtained by Kalman.... ..."

Cited by 293

### Table 2. Percentage of SIR steps: linear Gaussian model

2000

Cited by 293

### Table 1. Simulation results for a linear gaussian model.

### Table 2. Attributes in the Bach chorale data set. The key signature and time signature attributes were constant over the duration of the chorale. All attributes were treated as real numbers and modeled as linear-Gaussian observations (4a). Attribute Description Representation

1997

"... In PAGE 17: .... S. Bach apos;s Chorales, obtained from the UCI Repository for Machine Learning Databases (Merz amp; Murphy, 1996) and originally discussed in Conklin and Wit- ten (1995). Each event in the sequence was represented by six attributes, described in Table2 . Sixty-six chorales, with 40 or more events each, were divided into a training set (30 chorales) and a test set (36 chorales).... ..."

Cited by 279

### Table 2. Attributes in the Bach chorale data set. The key signature and time signature attributes were constant over the duration of the chorale. All attributes were treated as real numbers and modeled as linear-Gaussian observations (4a). Attribute Description Representation

1997

"... In PAGE 17: .... S. Bach apos;s Chorales, obtained from the UCI Repository for Machine Learning Databases (Merz amp; Murphy, 1996) and originally discussed in Conklin and Wit- ten (1995). Each event in the sequence was represented by six attributes, described in Table2 . Sixty-six chorales, with 40 or more events each, were divided into a training set (30 chorales) and a test set (36 chorales).... ..."

Cited by 279

### Table 2. Attributes in the Bach chorale data set. The key signature and time signature attributes were constant over the duration of the chorale. All attributes were treated as real numbers and modeled as linear-Gaussian observations (4a). Attribute Description Representation

"... In PAGE 17: .... S. Bach apos;s Chorales, obtained from the UCI Repository for Machine Learning Databases (Merz amp; Murphy, 1996) and originally discussed in Conklin and Wit- ten (1995). Each event in the sequence was represented by six attributes, described in Table2 . Sixty-six chorales, with 40 or more events each, were divided into a training set (30 chorales) and a test set (36 chorales).... ..."

### Table 2: Percentage of SIR steps: linear Gaussian model With N = 500 trajectories, the estimates obtained using MC methods are similar to those obtained by Kalman. The SIS algorithms have similar performances to the bootstrap lter for a smaller computational cost. The most interesting algorithm is based on the optimal importance function which limits seriously the number of resampling steps. 6.2. Nonlinear series. We consider here the following nonlinear reference model [7, 23, 35, 56]:

1998

Cited by 169