Next: Scenario 1. Single landmark Up: Real Experiments Previous: The Real Scenarios
We describe the experimentation carried out in each one of the four scenarios described above. We have used the parameters obtained through the Genetic Algorithm approach described in Chapter 5 (discarding those that are not used in the final version of the Navigation system). For each scenario, there is a brief discussion of the results. In each of these scenarios we have defined different starting points (two starting points in scenarios 1 and 2, and three in scenarios 3 and 4). We have run 40 trials for each starting point and stored the following statistics:
The relevant statistics of the experiments are shown in Table 6.2.
| Scenario | Success | #d.t. | Winning moving bids | Winning looking bids | |||||
| class | rate | TT | RE | PS | TT | RM | RE | PS | |
| 1 | 100% | 0 | 100% | 0% | 0% | 0% | 54% | 0% | 46% |
| 2 | 100% | 0 | 79% | 0% | 21% | 0% | 66% | 0% | 34% |
| 3 | 85% | 0 | 78% | 0% | 22% | 0% | 56% | 0% | 44% |
| 4 | 84% |
0: 24% 1: 58% 2: 18% |
67% | 2% | 31% | 3% | 41% | 0% | 56% |