Kernel Negative ε Dragging Linear Regression for Pattern Classification

Table 3

Accuracies (%) of different methods on the LFW database.

Number of training samples per class	6	7	8	9	10
Our method	35.95	38.00	39.13	40.27	41.10
DLSR	14.46	14.92	14.58	15.43	14.68
CLSR	34.45	35.93	35.15	36.65	35.27
NNLS	33.86	36.58	36.04	36.69	37.31
-SVM	32.88	34.76	35.70	36.52	38.59
KNN	20.34	21.28	21.12	21.26	21.53
SRC	34.15	36.70	38.06	39.31	39.00
LRC	32.12	35.16	34.78	36.50	40.05
NDLR	35.22	37.49	38.40	39.29	40.69