### Table 4: Algorithm comparisons on 2GM-2GM test data, with various prior assumptions for the SG algorithms.

"... In PAGE 30: ... The restricted \window sizes quot; for the SG3 and SG4 algorithms may be e ectively limiting the sample sizes and thus increasing the deviation of the sampled proportions from the expected proportions. This may explain the results, in Table4 , for the SG3 and SG4 algorithms with the equal prior assumption. 4.... ..."

### Table B.16: Inaccurate Assumptions for Prior and Re- sponse Function. True Prior and Reponse Function are Asymmetric and Weibull.

2006

### Table 5.1: Algorithm for scale mixture source separation with no assumptions made on priors

2007

### Table 1: Parameterization of the unrestricted likelihood and prior ** VAR under the assumption of trend stationarity where Y is a Tx3 matrix with typical row ( )

"... In PAGE 13: ... They are13 To maintain compatibility with DeJong and Whiteman (1994), and hence to emphasize the role consumption plays in this analysis, my specification of prior and likelihood functions closely follows theirs.14 Table1 summarizes the parameterization of the unrestricted likelihood function and the default settings of the unrestricted prior density of the parameters . The prior 11One can formulate similar hypotheses of interest for the constant discount rate model.... ..."

### Table 2: Separator symbols discovered in the Sher- lock Holmes corpus. Once one or more separator symbols have been found, traditional parsing techniques may be used to segment the text. Many data sequences simply will not have sepa- rator symbols. For example, a database may store elds in a le based on their bit length only. In such situations entropic chunking must be used if no prior assumptions about the structure of the data are to be made.

1998

"... In PAGE 2: ... Entropic chunking may be used to discover this automatically, by recording which symbols occur immediately prior to a large jump in entropy. Table2 lists, in decreasing order of frequency, sep- arator symbols discovered in the Sherlock Holmes corpus. The ^ symbol precedes the majority of sud- 80 Hutchens and Alder... ..."

Cited by 14

### Table 2: Separator symbols discovered in the Sher- lock Holmes corpus. Once one or more separator symbols have been found, traditional parsing techniques may be used to segment the text. Many data sequences simply will not have sepa- rator symbols. For example, a database may store elds in a le based on their bit length only. In such situations entropic chunking must be used if no prior assumptions about the structure of the data are to be made.

1998

"... In PAGE 2: ... Entropic chunking may be used to discover this automatically, by recording which symbols occur immediately prior to a large jump in entropy. Table2 lists, in decreasing order of frequency, sep- arator symbols discovered in the Sherlock Holmes... ..."

Cited by 14

### Table 5: Trade Parameter Estimates Under Alternative Prior Distributions

2001

"... In PAGE 3: ...able 4: Measures of Fit for Exports and Imports. ........................................ 28 Table5 : Trade Parameter Estimates Under Alternative Prior Distributions..... In PAGE 22: ... It is thus worthwhile to ask how alternative assumptions on prior distributions would influence parameter estimates. Table5 illustrates trade parameter estimates for the base case (prior distributions and estimates shown in Table 2) and two additional parameter priors. In Prior 1, support points are the same as in the base case except that the upper support point is reduced to six for the import elasticities and three for export elasticities.... In PAGE 22: ...9. Table5 also provides the first three moments for each of the three prior distributions. As is clear from Table 5, the choice of parameter prior distributions does influence the parameter estimates.... In PAGE 22: ...istribution is 0.9. Table 5 also provides the first three moments for each of the three prior distributions. As is clear from Table5 , the choice of parameter prior distributions does influence the parameter estimates. For both export and import elasticities, Prior 1 exhibits reduced variance and strongly reduced skewness relative to the base.... ..."

### Table 5. An example of the concatenation phase of algorithm 3 on n = 6 processors. respect to the number of message transmissions and the number of operations performed.A further advantage of algorithm 3 is that all processors will hold identical re- sults upon its completion, since each block bi is uniquely determined by processor pi prior to the concatenation phase. Algorithm 2 does not possess this property, in case numerical roundo errors violate the assumption of associativity (e.g., addition on binary computers is not necessarily associative).

in Performance Evaluation and Modeling of Reduction Operations on the IBM RS/6000 SP Parallel Computer

### Table 1: Models for Lag order 1 and LR Type Tests Assumption for Notation for

"... In PAGE 20: ... Speci cally we have LRPC trace(r0) d ! tr(DPC) and LRPC max(r0) d ! max(DPC), where DPC = Z 1 0 BPCdB0 n?r0 0 Z 1 0 BPCBPC0ds ?1 Z 1 0 BPCdB0 n?r0 and BPC is a trend adjusted version of Bn?r0. Critical values for these tests are given in Perron amp; Campbell (1993, Table1 ) (see also Rahbek (1994, Table 1)). It should be noted that the null distribution is unchanged if in fact there is no trend and intercept term in the true DGP ( 0 = 1 = 0).... In PAGE 21: ... Obviously, this distribution is free of unknown nuisance parameters and does not depend on the actual trending properties of the process. Critical values are given in L utkepohl amp; Saikkonen (1997b, Table1 ) and Saikkonen amp; L utkepohl (1997, Table 1). Another estimator of the mean and trend parameters proposed by Saikkonen amp; L utkepohl (1997) is based on a GLS approach applied to the model yt = 0 + 1t + xt.... In PAGE 29: ... 3.5 Summary of Models and Test Statistics A summary of the model types in a VAR(1) setting and the corresponding LR type test sta- tistics is given in Table1 . The associated asymptotic distributions under local alternatives are given in Table 2.... In PAGE 43: ... The asymptotic null distributions of these statistics are functionals of Brownian motions and depend on n ? r0 and the trending properties of the process. Some critical values for the case without deterministic trend ( 1 = 0) are given in Table1 of Bewley amp; Yang (1995)2. Critical values of the test statistics BY i trace and BY i min for the case 1 6 = 0, 0 1 = 0, may be found in Table 1 of Yang amp; Bewley (1996).... In PAGE 43: ... Some critical values for the case without deterministic trend ( 1 = 0) are given in Table 1 of Bewley amp; Yang (1995)2. Critical values of the test statistics BY i trace and BY i min for the case 1 6 = 0, 0 1 = 0, may be found in Table1 of Yang amp; Bewley (1996). In the latter article the authors also note that for a given cointegrating rank r0, the pair of hypotheses H0 : 1 = 0 vs.... In PAGE 43: ...5) where H0 is rejected if BYmax(r0) is small. Critical values appear in Table1 of their article. Bewley amp; Yang (1995) also propose two other tests for the case 1 = 0.... In PAGE 45: ... The asymptotic null distribution of this test statistic is nonstandard. Its critical values are tabulated in Table1 of Stock amp; Watson (1988). This test may be regarded as nonparametric because the VAR(p) may just be an ap- proximation to a more general DGP of the common trends t.... ..."

### Table 2. Under the large-sample assumption, the method of

"... In PAGE 6: ... We transformed the #28continuous#29 data from the Lipid study to the binary variables D and Y using the same method as Balke and Pearl #281994#29. The resulting data set is shown in Table2 . Using the large-sample assumption, Balke and Pearl #281994#29 use the given data to derive the bounds 0:39 #14 ACE#28D ! Y #29 #14 0:78.... In PAGE 7: ...Table2 : Observed data for the Lipid study and the Vitamin A study Lipid Study Vitamin A Study z d y Observations Observations 0 0 0 158 74 0 0 1 14 11514 0 1 0 0 0 0 1 1 0 0 1 0 0 52 34 1 0 1 12 2385 1 1 0 23 12 1 1 1 78 9663 parameters#29 and the density of Figure 6b corresponds to skewed priors. Rather remarkably, even with only 337 cases in the data, both posterior distributions are highly concentrated within the large-sample bounds.... ..."