An Efficient Normalized Rank Based SVM for Room Level Indoor WiFi Localization with Diverse Devices

Table 4

Algorithms comparison based on the percentage of the well-recognized positions; the radio map is defined upon the proposed normalized rank transformation.

Device type	Classification type
	SVM	Artificial Neural Network	Naïve Bayes	Random Forest	KNN ()
	L2-R
Oppo A31c
BBK	98.75%
Coolpad 8730L	100%	100%	95.83%
Gionee	100%	93.54%
HTC One E8	98.75%	98.54%	96.25%
Huawei GRA-CL00	100%	98.54%
Lenovo A788t	99.79%	96.67%
Meizu	100%	100%	100%	93.54%	99.37%
Oppo R7c	99.58%	93.54%
Xiaomi	99.17%
Xiaomi Cancro	100%	93.54%
Samsung klteduoszn	100%	99.79%	99.58%
Meizu M2 note	100%	98.75%	99.58%		92.50%
Samsung trlteduosctc	99.58%	99.17%	99.79%
BBK Vivo	100%	96.46%	69.38%	63.96%	70.83%
Standard device accuracy	100%	100%	100%		90.83%
Recall (ratio)	1	1	1	0.89	0.91
Precision (ratio)	1	1	1	0.94	0.94

We put in bold the percentages lower than 90% for easiness of observation.