### Table 3: Confusion matrix for engagement detection on independent test data

### Table 3: Confusion matrix for engagement detection on independent test data

### Table 3-6ias, standard deviation, and mean absolute value for each of the equations based on the independent data

in Proceedings:

"... In PAGE 84: ... A third test with 1992 bulk collections from Brunswick and Jones Counties, North Carolina, suggested that 4 weeks was optimum in terms of germination rate and also for total germination for one of the lots (table 4). In the fourth stratification test, there were Table3 -Results of stratification tests on old stored seeds from the NC Division of Forest Resources and the 1992 bulk collections from Milltail Creek and Sycamore Road stands in Dare County, N.C.... ..."

### Table 22. Table 22 Independent Variables for Hypothesis 4 Independent Variable Data Type Variable

2003

"... In PAGE 94: ...83 Table22 -continued Independent Variable Data Type Variable Name total number of Ph.D.... In PAGE 96: ...3104 (that is, the two independent variables explain 31% of the variation in the dependent variable). The p-value ( Table22 ) for the regression was lt;0.001 which is much less than the 0.... ..."

### Table 7. Application/ Success/ Independence Frequencies for Contrived Data

"... In PAGE 9: ... The statistics are shown in Table 6. Table7 is similar to Table 5 except that it is for con- trived data. In this case, we clearly find that the GCD test and the Banerjee test are less exact than the Power test and the K test.... ..."

### Table 5. Application/Success/Independence Frequencies for Practical Data

"... In PAGE 8: ... Table 4 lists the number of lines, subroutines and array pairs tested. Table5 shows the usage and success frequencies of the dependence tests for the two packages. The notations are similar to those in Goff et al.... In PAGE 9: ... The statistics are shown in Table 6. Table 7 is similar to Table5 except that it is for con- trived data. In this case, we clearly find that the GCD test and the Banerjee test are less exact than the Power test and the K test.... ..."

### TABLE 1. DATA INDEPENDENT CONNECTIONS OF AS-VO COMPUTATION GRAPH.

2004

Cited by 2

### Table 3. Inversion results using independent SAR and GPS data.

### Table 5: Comparison of independent model with ob- served data (by case)

"... In PAGE 6: ... PV = C2 3C1 + C2 (3) The data from table 2 yields an estimate of PV = 0:0958 for the probability of activation of an indepen- dent fault in a 3-version con guration. Table5 com- pares the probability of activation of 1, 2 and 3 faults as predicted by a model assuming independence be- tween versions, with the observed values. The ob- served probability of two simultaneous errors is lower than predicted by the independent model, while the observed probability of three simultaneous errors is higher than predicted.... ..."

### Table 5: Comparison of independent model with ob- served data (by case)

"... In PAGE 6: ... PV = C2 3C1 + C2 (3) The data from table 2 yields an estimate of PV = 0:0958 for the probability of activation of an indepen- dent fault in a 3-version con guration. Table5 com- pares the probability of activation of 1, 2 and 3 faults as predicted by a model assuming independence be- tween versions, with the observed values. The ob- served probability of two simultaneous errors is lower than predicted by the independent model, while the observed probability of three simultaneous errors is higher than predicted.... ..."