Data Anomaly Detection through Semisupervised Learning Aided by Customised Data Augmentation Techniques

Table 4

DA composition strategies in this study.

Data pattern	Normal	Missing	Minor	Outlier	Square	Trend	Drift	Total
No DA	13,575	2,942	1,775	527	2,996	5,778	679	28,272
Aug_1	—	—	10,650	3,162	—	—	4,074	46,158
Aug_2	—	—	10,650	6,324	—	—	4,074	49,320
Aug_3	—	—	10,650	12,648	—	—	4,074	55,644
Aug_4	—	—	10,650	9,486	—	—	8,148	56,556
Aug_5	—	—	10,650	12,648	—	—	8,148	59,718
Aug_6	—	—	10,650	12,648	—	—	12,222	63,792
Aug_7	—	—	10,650	12,648	—	11,562	12,222	75,354
Aug_8	—	—	10,650	15,810	—	—	16,296	82,590
Aug_9	—	—	10,650	18,972	—	—	20,370	89,826