### Table 1: ATSP Results

"... In PAGE 3: ... Dayan (1992)] update rule run in an online mode (update preference matrices, right after new solutions are generated) as applied to the ATSP, and a Monte Carlo update rule for the QAP. Table1 shows the results on some instances of the ATSP from the TSPLIB [Reinelt (1991)] benchmark set using PBIL2 update [Baluja (1994)] in offline settings, with parameter l=10 and PBIL positive and negative learning rates equal 0.01.... In PAGE 3: ...ositive and negative learning rates equal 0.01. The rest of algorithm was the same as used in the RL context [Miagkikh amp;Punch(1999b)]. The meaning of the columns of Table1 are as follows: Opt./BKS.... ..."

### Table 1* Back Propagation Genetic Algorithm

1995

"... In PAGE 11: ... In each case network architecture is that of Figure 1. The parameters used in the training of the network are given in Table1 . Two measures of fit, namely the sum of squares deviation (Ssq.... In PAGE 17: ... Table1 . The parameter settings to run the neural network using both the back propagation and the genetic algorithms to run all three problems are given.... ..."

### Table 1. Comparisons of results over ten seeds

1998

"... In PAGE 6: ... The 600,000 value was appropriate for the larger Armour and Buffa problem while the smaller Bazaraa problem converged in much fewer number of solutions searched. Objective function values from the perimeter metric are in Table1 , where the best, median, worst and standard deviation over ten random seeds are shown. The twenty department Armour and Buffa (A amp;B) problem was studied with maximum aspect ratios of 10, 7, 5, 4, 3 and 2, which represent problems ranging from lightly constrained to extremely constrained.... ..."

Cited by 3

### Table 4: Genetic Algorithm Performance on External Simulator

1999

"... In PAGE 34: ... PLACE TABLE 4 HERE We next looked at the 50 best solutions and 50 randomly selected so- lutions in the hybrid genetic algorithm population at 60, 70, 80 and 100 thousand evaluations using the external simulator. The results presented in Table4 suggest the solutions found at 60 thousand evaluations are basically as good as the solutions at 100 thousand evaluations. The best solution ever seen on this problem as evaluated by the external simulator is found in the randomly selected set after 70,000 evaluations: a mean time at dock of 378.... ..."

Cited by 10

### Table 4: Genetic Algorithm Performance on External Simulator

1999

"... In PAGE 34: ... PLACE TABLE 4 HERE We next looked at the 50 best solutions and 50 randomly selected so- lutions in the hybrid genetic algorithm population at 60, 70, 80 and 100 thousand evaluations using the external simulator. The results presented in Table4 suggest the solutions found at 60 thousand evaluations are basically as good as the solutions at 100 thousand evaluations. The best solution ever seen on this problem as evaluated by the external simulator is found in the randomly selected set after 70,000 evaluations: a mean time at dock of 378.... ..."

Cited by 10

### Table 2: Results for selected distance metrics on two clustering problems.

2003

Cited by 136

### Table 2: Results for selected distance metrics on two clustering problems.

2003

Cited by 136

### Table 2: Results for selected distance metrics on two clustering problems.

2003

Cited by 136

### Table 2: Results for selected distance metrics on two clustering problems.

2003

Cited by 136

### Table 7: Recommended distance metrics. This table matches distance metrics with local planners. The labels associ- ated with the distance metrics are as shown at the bottom of the table. The label in bold is the metric that was best for that local planner in our tests.

1998

"... In PAGE 25: ... 5 Conclusion The main goal of our study was to determine good combinations of distance metrics and local planners for use by PRMs in cluttered environments. Our results, as reported in Table7 , include recommendations for selecting distance metrics for various local planners in different types of environments. Generally, a good choice is the Scaled Euclidean metric, where more weight is placed on the position coordinates as the environment becomes more cluttered.... ..."

Cited by 59