Research Article
Improving Accuracy and Simplifying Training in Fingerprinting-Based Indoor Location Algorithms at Room Level
Table 7
Relative accuracy by algorithm based on the number of WAPs when training with border locations (part III).
| #WAPs | Random Tree | REP Tree | SVM | NN
| NBC |
| 40 | 91.50 ± 3.76 | 92.50 ± 3.59 | 98.00 ± 1.00 | 97.00 ± 1.63 | 87.17 ± 1.83 |
| 5 | 88.33 ± 1.29 | 82.50 ± 4.17 | 68.17 ± 1.74 | 80.00 ± 2.98 | 80.50 ± 1.07 |
| 8 | 87.83 ± 1.83 | 92.17 ± 2.89 | 87.50 ± 2.71 | 86.33 ± 1.94 | 81.17 ± 1.07 |
| 10 | 97.33 ± 1.53 | 93.33 ± 3.07 | 98.17 ± 0.50 | 92.17 ± 1.30 | 97.00 ± 0.67 |
| 16 | 95.67 ± 3.00 | 92.83 ± 3.66 | 98.33 ± 0.75 | 95.17 ± 1.89 | 96.67 ± 1.05 |
| 20 | 93.83 ± 3.66 | 93.83 ± 3.95 | 97.50 ± 0.83 | 94.67 ± 1.63 | 95.50 ± 1.30 |
| 32 | 95.83 ± 4.43 | 92.83 ± 3.66 | 99.33 ± 0.82 | 98.50 ± 0.90 | 87.50 ± 1.54 |
|
|