### Table 2. Observation Sequences

"... In PAGE 5: ... We used a set of 9 { 10 CVF wavelengths and 1 { 2 narrowband lters toward each nebula, in four sequences. Each sequence contains observations at three to four wavelengths, as listed in Table2 . We observed all four sequences for each nebula, except for vdB 135, where no Sequence 3 observations were made.... ..."

### Table 1. Observing Sequence

"... In PAGE 6: ...2. Observations The measurements consisted of the series of skydips and spectral measurements listed in Table1 . Each sequence took two hours to complete.... ..."

### Table 1. Observing Sequence

"... In PAGE 6: ...2. Observations The measurements consisted of the series of skydips and spectral measurements listed in Table1 . Each sequence took two hours to complete.... ..."

### Table 1. A toy example of an observation sequence for a region of DNA at an ORF boundary

"... In PAGE 2: ... These three sequences are aligned by associating each codon usage and expression observation with a specific position in the DNA sequence. Table1 provides a toy example illustrating each these sequences. For the analysis considered here, we assume that we are given the (predicted) coordinates of every ORF in the genome.... ..."

### Table 1. Atoy example of an observation sequence for a region of DNA at an ORF boundary

2003

"... In PAGE 2: ... These three sequences are aligned by associating each codon usage and expression observation with a specific position in the DNA sequence. Table1 provides a toy example illustrating each these sequences. Forthe analysis considered here, we assume that we are given the (predicted) coordinates of every ORF in the genome.... ..."

### TABLE II A PROBLEM WITH TWO POSSIBLE SEQUENCES OF OBSERVATIONS.

2003

Cited by 2

### Table 2. An example transformation from a sequence to a set of observations

### Table 2. The results of predicting next viewing change as a function of difierent number of states (S) and observations (O) for two training sets and two test sets. Values correspond to probabilities of correctly predicting the observation sequences given the model and the Viterbi MAP solutions. The right column shows chance performance levels. Average length of the observation sequences was 27.

2004

"... In PAGE 5: ...The numbers in Table2 refer to the average probabilities of correctly predict- ing the observation sequences given the model and the Viterbi MAP solution over 100 Monte Carlo trials. This reduces to sampling from the state-dependent obser- vation vectors given the Viterbi-predicted state for each observation value.... ..."

Cited by 2