Imbalanced Learning Based on Logistic Discrimination

Table 5

The -measures and standard errors of ILLD, LD, LD-US, and LD-OS.

Data set	ILLD	LD-US	LD-OS	LD
d1	0.6446 ± 0.1734	0.3955 ± 0.1161^•	0.6587 ± 0.1788	0.7322 ± 0.1060
d2	0.6335 ± 0.0855	0.4778 ± 0.0883^•	0.5826 ± 0.1003	0.6062 ± 0.1561
d3	0.5207 ± 0.1391	0.3884 ± 0.0974^•	0.5475 ± 0.1507	0.5019 ± 0.1680
d4	0.9532 ± 0.0046	0.9170 ± 0.0416^•	0.9281 ± 0.0130^•	0.9042 ± 0.0221^•
d5	0.5204 ± 0.0353	0.4353 ± 0.0302^•	0.4546 ± 0.0399^•	0.4025 ± 0.0478^•
d6	0.8047 ± 0.0367	0.3926 ± 0.0757^•	0.6637 ± 0.0888^•	0.8149 ± 0.0655
d7	0.2975 ± 0.0961	0.1764 ± 0.0681^•	0.2704 ± 0.1140	0.2391 ± 0.1156
d8	0.5621 ± 0.1615	0.4358 ± 0.1162	0.4809 ± 0.1501	0.4170 ± 0.1798
d9	0.5825 ± 0.0667	0.4812 ± 0.0724^•	0.4966 ± 0.0936^•	0.5193 ± 0.1135
d10	0.7407 ± 0.0348	0.6396 ± 0.0387^•	0.6833 ± 0.0245^•	0.6999 ± 0.0412^•
d11	0.7853 ± 0.0378	0.7314 ± 0.0433^•	0.7392 ± 0.0528	0.7526 ± 0.0604
d12	0.1837 ± 0.0186	0.1358 ± 0.0363^•	0.1487 ± 0.0282^•	0.1242 ± 0.0616^•
d13	0.8849 ± 0.0188	0.8817 ± 0.0130	0.9259 ± 0.0050°	0.8081 ± 0.0402^•
d14	0.6157 ± 0.0880	0.0475 ± 0.0133^•	0.0988 ± 0.0450^•	0.5455 ± 0.0825^•
Average	0.6235	0.4669	0.5485	0.5763