### Table 2: constant number of users/roles, varying permissions

2006

"... In PAGE 7: ...Table 2: constant number of users/roles, varying permissions In the rst set of experiments, we kept the number of users and roles constant, while changing the number of per- missions (and correspondingly, the number of permissions per role). Table2 describes the test parameters. Figure 3(b) shows the time taken by the algorithm, while Figure 3(a) shows the accuracy of the algorithm.... ..."

Cited by 3

### Table 3: Constant number of permissions/roles, varying users

2006

"... In PAGE 7: ... In the second set of experiments, we kept the number of roles and permissions constant while varying the number of users. Table3 describes the test parameters. Figure 4(a) shows the time taken by the algorithm, while Figure 4(b) shows the accuracy of the algorithm.... ..."

Cited by 3

### Table 4: Constant number of permissions, varying users/roles

2006

"... In PAGE 7: ... In the third set of experiments we keep the permissions constant, and vary the number of users as well as the num- ber of roles. Table4 describes the test parameters. Figure 5(a) shows the time taken by the algorithm, while Figure 5(b) shows the accuracy of the algorithm.... ..."

Cited by 3

### Table 2. Learning rate constants and number of time steps required for learning

"... In PAGE 4: ... Time steps required for learning as a function of time step duration, Dt 26 LIST OF TABLES Table 1. Comparison of several algorithms 17 Table2 . Learning rate constants and number of time steps required for learning 28 Table 3.... In PAGE 31: ... Due to the time required for the simulations, the last two data points for Q learning were found with averages over 10 systems rather than 100. Table2 shows the learning rates used for each of the three functions for both of the algorithms for various noise levels, and Table 3 shows learning rates for various time step durations. For the simulations described... ..."

### TABLE IX CHORD HIT RATE WITH CONSTANT NUMBER OF CLIENTS

2006

Cited by 1

### TABLE 2. CPU times and number of constants discovered.

### Table 13: Percentage of completely malicious tunnels (figure 27) constant number of com- number of com- false percentage of com-

2002

Cited by 53

### Table 3. Optimal learning rate constants and number of time steps required for learning

"... In PAGE 4: ... Comparison of several algorithms 17 Table 2. Learning rate constants and number of time steps required for learning 28 Table3 . Optimal learning rate constants and number of time steps required for learning... In PAGE 31: ... Due to the time required for the simulations, the last two data points for Q learning were found with averages over 10 systems rather than 100. Table 2 shows the learning rates used for each of the three functions for both of the algorithms for various noise levels, and Table3 shows learning rates for various time step durations. For the simulations described... ..."

### Table 2; The recognition rate and time for the original and modified recognizers with a constant number of templates per category.

"... In PAGE 4: ... The optimal number of templates per category was found out by experiments and for Japanese characters it was set to 40. In Table2 we can compare the recognition rate and computation time for the original and modified classifiers. Numbers of templates vectors per category are 10, 20 and 40.... ..."