seems to be a common misunderstanding that ML systems are constantly being trained on new data.. does anyone do online learning in practice?pic.twitter.com/gvLitptKLs
for some verticals it is standard practice to do online learning in prod, for instance with fraud detection (ad click fraud, banking)
-
-
some targeting models & recommendation engines would also use online learning, albeit the smaller ones are still doing batch training
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
great point -- makes me wonder when we will need to do online learning with images. facebook profiles generated by GANs?
-
online learning makes most sense when the labels are part of the data being received & when systems need to respond to fast trends.
- Show replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.