User Authentication Method Based on MKL for Keystroke and Mouse Behavioral Feature Fusion

Table 1

Summary of the related researches by using a combination of KD and MD.

Ref.	Year	Users	Method	Performance
[6]	2005	22	NN	FAR of 1.312% and FRR of 0.651%
[31]	2007	61	DT and SVM	Error rate of 1.5%
[32]	2009	20	Statistical analysis, feed-forward network with back-propagation, and k-nearest neighbour (k-NN)	Accuracy of 82.22–96.4%
[33]	2012	24	Bayesian network (BN)	EER of 8.21%
[34]	2014	31	BN, DT, and SVM	FAR of 2.10% and FRR of 2.24%
[5]	2015	67	Naive Bayes (NB) and SVM	FAR of 0.1% and FRR of 0.2%
[7]	2016	25	DT, Counter-Propagation Artificial Neural Network (CPANN), Artificial Neural Network (ANN), and SVM	Accuracy of 62.2%
[35]	2017	53	ANN, CPANN, and SVM	FRR of 5.7% and FAR of 0.1%
[36]	2018	41	RF, SVM, DT, and BN	Accuracy of 80.6%