Smartwatch-Based Legitimate User Identification for Cloud-Based Secure Services

Table 1

Average accuracy for legitimate user identification with 400 samples per window.

Sensor	Activities	Classifiers
		DT	SVM	KNN	NB
Acc.	Walking	0.9868 ± 0.0094	0.9391 ± 0.0181	0.9375 ± 0.0171	0.7616 ± 0.0201
	Walking up	0.8753 ± 0.0327	0.8005 ± 0.0422	0.7874 ± 0.0388	0.5876 ± 0.0540
	Walking down	0.8835 ± 0.0383	0.8447 ± 0.0534	0.7831 ± 0.0598	0.5619 ± 0.0909
	Running	0.8168 ± 0.0363	0.6381 ± 0.0554	0.5969 ± 0.0477	0.5311 ± 0.0325
	Jogging	0.9572 ± 0.0147	0.8816 ± 0.0163	0.9260 ± 0.0236	0.6234 ± 0.0384
Gyr.	Walking	0.9638 ± 0.0115	0.9177 ± 0.0268	0.9504 ± 0.0066	0.5149 ± 0.0367
	Walking up	0.8575 ± 0.0389	0.8362 ± 0.0333	0.8313 ± 0.0181	0.5423 ± 0.0683
	Walking down	0.8983 ± 0.0453	0.8062 ± 0.0339	0.8126 ± 0.0225	0.5476 ± 0.0595
	Running	0.8521 ± 0.0651	0.6737 ± 0.0390	0.7359 ± 0.0220	0.5866 ± 0.0483
	Jogging	0.9424 ± 0.0163	0.8964 ± 0.0272	0.8903 ± 0.0095	0.6629 ± 0.0376
Mag.	Walking	0.9193 ± 0.0208	0.7699 ± 0.0259	0.8551 ± 0.0339	0.6777 ± 0.0339
	Walking up	0.7600 ± 0.0549	0.5951 ± 0.0486	0.7107 ± 0.0405	0.5907 ± 0.0674
	Walking down	0.8298 ± 0.0543	0.7233 ± 0.0486	0.7055 ± 0.0671	0.5826 ± 0.0716
	Running	0.6384 ± 0.0611	0.5768 ± 0.0339	0.5916 ± 0.0481	0.3524 ± 0.0589
	Jogging	0.7056 ± 0.0461	0.4637 ± 0.0419	0.5873 ± 0.0361	0.3340 ± 0.0287