### Table 1.|SNP Estimation { Maximum Likelihood Surface.

"... In PAGE 10: ... Thus, for a given SNP speci cation, the smaller the R-squared of the regressions, the better the SNP model approximates the true density. In Table1 we present the maximum likelihood surface for three key models: (1) the basic ARCH(1) model, which is SNP(11100); (2) the ARCH model with many lags in the variance (given by SNP(1h100), which has 17 lags in the variance) to approx- imate the GARCH(1,1) speci cation; and (3) the preferred model from our selection procedure, which is SNP(1c121). The preferred model is a general nonlinear process with heterogeneous innovations.... In PAGE 10: ...ith 816 observations for each of three series implying a saturation ratio of 26.3.9 Table 1 indicates that the preferred model performs substantially better than the other two models according to all three model selection criteria. (Insert Table1 here.) The superior performance of the preferred model in matching the data is also re-... ..."

### Table 1: Posterior distribution of k for the 3 data sets based on a mixture model with random and default parameter values.

1994

"... In PAGE 10: ... In all the runs, the number of components never exceeded 24, hence the chosen value of kmax was inconsequential. Estimated posterior probabilities are given in Table1 . In each of the data sets, it is immediately apparent that there are a number of competing explanations of the data which are tenable.... In PAGE 18: ... Similar plots were obtained for the other data sets. Proportions of accepted `split or combine apos; moves vary between 8% and 14% ( Table1 ). For dimension changing moves, these proportions are satisfactory and show that our proposal based on adjacency is sensible.... ..."

Cited by 2

### Table 1: Posterior distribution of k for the 3 data sets based on a mixture model with random and default parameter values.

1994

"... In PAGE 10: ... In all the runs, the number of components never exceeded 24, hence the chosen value of kmax was inconsequential. Estimated posterior probabilities are given in Table1 . In each of the data sets, it is immediately apparent that there are a number of competing explanations of the data which are tenable.... In PAGE 18: ... Similar plots were obtained for the other data sets. Proportions of accepted `split or combine apos; moves vary between 8% and 14% ( Table1 ). For dimension changing moves, these proportions are satisfactory and show that our proposal based on adjacency is sensible.... ..."

Cited by 2

### Table 1: Posterior distribution of k for the 3 data sets based on a mixture model with random and default parameter values.

1994

"... In PAGE 10: ... In all the runs, the number of components never exceeded 24, hence the chosen value of kmax was inconsequential. Estimated posterior probabilities are given in Table1 . In each of the data sets, it is immediately apparent that there are a number of competing explanations of the data which are tenable.... In PAGE 18: ... Similar plots were obtained for the other data sets. Proportions of accepted `split or combine apos; moves vary between 8% and 14% ( Table1 ). For dimension changing moves, these proportions are satisfactory and show that our proposal based on adjacency is sensible.... ..."

Cited by 2

1994

"... In PAGE 10: ...Enzyme data : R = 2:86, = 1:45, = 0:122, = 2, g = 0:2, h = 1:22, = 1 Acidity data : R = 4:18, = 5:02, = 0:057, = 2, g = 0:2, h = 0:573, = 1 Galaxy data : R = 25:11, = 21:73, = 0:0016, = 2, g = 0:2, h = 0:016, = 1 Estimated posterior probabilities are given in Table1 . In each of the data sets, it is immediately apparent that there are a number of competing explanations of the data which are tenable.... In PAGE 16: ... Similar plots were obtained for the other data sets. Proportions of accepted `split or combine apos; moves vary between 8% and 14% ( Table1 ). For dimension changing moves, these proportions are satisfactory and show that our proposal based on adjacency is sensible.... ..."

Cited by 2

### Table 2. Patient Demographics and Comparison of Diagnostic Methods

"... In PAGE 4: ... This analysis clustered the test set into three main groups: one malignant group (n H11005 10), one benign group (n H11005 7), and one indeterminate group (two FA, two FVPTC, and one HYP). These results were compared with the preoperative FNA diagnosis and the final histological diagnosis of the resected specimens ( Table2 ). Sixteen of the 17 cases predicted to be either benign or malignant by microarray analysis were confirmed by pathological diagnosis.... In PAGE 5: ...histological diagnoses of these five indeterminate cases were two FA, two FVPTC, and one HYP. Interestingly, all five cases were deemed suspicious on preoperative FNA ( Table2 ). The histological sections of these five cases were reviewed (by T.... In PAGE 5: ... In comparison to FNA diagnosis by cytology, the microarray data showed higher concordance rate to the final tissue diagnosis. As is summarized in Table2 , the classification of benign or malignant by microarray analysis of FNA was 100% con- cordant to the histological diagnosis versus 76% concor- dance with the preoperative cytological FNA diagnosis. Eight of twenty-two FNAs were deemed indeterminate, including follicular and suspicious for PTC lesions, on preoperative cytology.... ..."

### Table 2: Comparison of the two Methods for bias estimation.

2002

"... In PAGE 10: ... Thus, we can also use the estimated Hurst exponent to obtain an estimate of the bias CQ. Comparison of the two methods Table2 compares the two model fitting algorithms using four data sets gener- ated with the CQ-model and bias ranging from 0.... ..."

Cited by 27

### Table 6 : Comparison of Complex Surface Methods

"... In PAGE 5: ... 43 Table 3: Determining Silhouettes. 47 Table 4: Computational Requirements 54 Table 5: Display Rate Requirements 55 Table6 : Comparison of Complex Surface Methods 56 ... ..."

### Table 2: Comparison of the two Methods for bias estimation.

"... In PAGE 10: ... Thus, we can also use the estimated Hurst exponent to obtain an estimate of the bias a0 . Comparison of the two methods Table2 compares the two model fitting algorithms using four data sets gener- ated with the a0 -model and bias ranging from 0.... ..."

### Table 2: Comparison of the two Methods for bias estimation.

"... In PAGE 10: ... Thus, we can also use the estimated Hurst exponent to obtain an estimate of the bias CQ. Comparison of the two methods Table2 compares the two model fitting algorithms using four data sets gener- ated with the CQ-model and bias ranging from 0.... ..."