Random Forest Based Coarse Locating and KPCA Feature Extraction for Indoor Positioning System

Table 2

Performance of SVM and RF for classification.

Support vector machine			Random forest
Regions	Number of RPs	Locating correctly	Regions	Number of RPs	Locating correctly
1	422	401	1	422	419
2	53	35	2	53	49
3	141	137	3	141	140
4	83	79	4	83	82
5	129	109	5	129	129
Coarse locating accuracy: 91.9%			Coarse locating accuracy: 98.9%