HIGHLIGHTS
SUMMARY
The representative images seem to have different characteristics for each class, and similar images are close to their representative images. The image of class in Fig 3 shows that the classifier recognises the same class even if the data are shifted in the time direction. Considering the classes (classes ("1080 Lines"), ("Repeating_Blips"), ("Chirp"), ("Helix"), and_(24) ("Scratchy")), unsupervised learning classifies the Gravity Spy labels (noted parentheses) as one class. A previous s tudy17 on supervised learning with the Gravity Spy labels indicated the existence of a subclass that might be in the "Scattered_Light" class . . .
If you want to have access to all the content you need to log in!
Thanks :)
If you don't have an account, you can create one here.