A Folded Concave Penalty Regularized Subspace Clustering Method to Integrate Affinity and Clustering

Table 2

Performance on ORL.

Methods	SSC	LRR	LRSC	KSSC	SSC-OMP	EDSC	S^3C	SSRSC	TLRR	KTRR	Our method
CA (%)
	77.6	74.9	68.5	76.2	78.4	82.9	81.2	78.2	76.8	87.5		88.3
	78.5	73.8	67.3	75.6	77.5	81.6	81.6	79.1	75.4	89.0		89.3
	79.1	73.8	65.4	73.0	74.7	79.1	82.8	79.9	74.7	90.4		90.8
NMI
	0.7711	0.8376	0.7239	0.7434	0.7676	0.8239	0.8410	0.8118	0.7965	0.8428		0.8656
	0.8271	0.8197	0.7094	0.7832	0.8156	0.8053	0.8408	0.8313	0.7889	0.8804		0.9056
	0.8355	0.8372	0.7323	0.8184	0.8382	0.8338	0.8663	0.8359	0.7726	0.9038		0.9210
Purity
	0.7520	0.7860	0.7364	0.7455	0.7564	0.7993	0.8100	0.8020	0.7913	0.8330		0.8450
	0.7610	0.7620	0.7049	0.7385	0.75636	0.7564	0.7785	0.7700	0.7311	0.8355		0.8444
	0.7483	0.7640	0.6687	0.7448	0.7411	0.7563	0.7937	0.7610	0.6773	0.8500		0.8665