Concept Drift Datasets for - LSTM-SCCM: Long Short-Term Memory Stream Cruise Control Method for Automated Drift Adaptation

Datasets Guide: Drift Locations: Abrupt marks sudden drift onset. Incremental drift recurs every $K$ points. Gradual shows drift as a sequence of concepts c1 and c2. E.D.: Euclidean distance between coefficients in space. C-Start: Start Concept. C-End: End Concept. E.D. (Consec.): E.D. between each two consecutive concepts in sequence. Dataset Drift Type Data Points Dimensions Noise Drift Locations E.D(C-Startto C-End) E.D.(Consec.) DS1 Abrupt 1k 2 10 500 128.05 128.05 DS2 Abrupt 10k 10 20 5k 327.23 327.23 DS3 Abrupt 20k 50 30 10k 344.95 344.95 DS4 Abrupt 100k 500 50 50k 324.11 324.11 DS5 Incremental 1k 2 10 every 100 399.93 [4.9, 6.1, 7.9,10.5, 14.8, 22.2, 37.0, 74.0, 222.1] DS6 Incremental 10k 10 20 every 1k 1205.10 [14.8, 18.5, 23.9, 31.8, 44.6, 66.9, 111.5, 223.1, 669.5] DS7 Incremental 20k 50 30 every 2k 1439.39 [17.7, 22.2, 28.5, 38.0, 53.3, 79.9, 133.2, 266.5, 799.6] DS8 Incremental 100k 500 50 every 10k 1671.75 [20.6, 25.7, 33.1, 44.2, 61.9, 92.8, 154.7, 309.5] DS9 Gradual 1k 2 10 [250-c1, 100-c2, 100-c1, 200-c2, 100-c1, 250-c2] 601.92 [377.4, 399.9, 829.2, 829.2, 624.36] DS10 Gradual 10k 10 20 [2.5k-c1, 1k-c2, 1k-c1, 2k-c2, 1k-c1, 2.5k-c2] 2049.14 [1158.3, 1205.1, 1885.4, 1885.4, 2167.7] DS11 Gradual 20k 50 30 [5k-c1, 2k-c2, 2k-c1, 4k-c2, 2k-c1, 5k-c2] 1756.43 [1469.9, 1439.3, 1875.3, 1875.3, 1843.6] DS12 Gradual 100k 500 50 [25k-c1, 10k-c2, 10k-c1, 20k-c2, 10k-c1, 25k-c2] 1364.12 [2126.2, 2032.1, 1238.2, 1238.2, 1322.0]