Research Article
Acoustic Signal NLOS Identification Method Based on Swarm Intelligence Optimization SVM for Indoor Acoustic Localization
Table 1
Accuracy comparisons of each classifier.
| | Algorithm | Number of feature dimension when accuracy is highest | Accuracy |
| | SVM (RBF as kernel function) | 5 | 81.01% | | SVM (linear kernel function) | 9 | 80.91% | | SVM (polynomial kernel function) | 8 | 77.12% | | SVM (sigmoid as kernel function) | 9 | 51.50% | | Logistic regression | 6 | 83.33% | | LDA | 7 | 81.34% | | Naive Bayes | 9 | 80.02% | | GridSearch+SVM | 9 | 73.80% | | GA+SVM | 9 | 82.22% | | PSO+SVM | 7 | 80.22% | | AFSA+SVM | 6 | 75.00% | | Advanced GA+SVM | 9 | 93.26% | | Advanced PSO+SVM | 9 | 97.12% |
|
|
