User Controls
Machine learning to permaban Scrotardo
-
2017-06-30 at 3:49 AM UTC
-
2017-06-30 at 3:50 AM UTCEnough of Spamnaldo J AIDS please
-
2017-06-30 at 3:50 AM UTCWe need an ignore function so that scron becomes dickless. We can just hit 'ignore' and boom.
-
2017-06-30 at 3:51 AM UTC
-
2017-06-30 at 4:02 AM UTCOr we could kill him. Who's closest?
-
2017-06-30 at 4:06 AM UTC
-
2017-06-30 at 5:30 AM UTC
Originally posted by Scrawny I yeah but I already know you are coming and I dont want to be alive anyways. Killing me would be the chase of your life and you would go missing and vanish suddenly. like price is right says… COME ON DOWN!!!!! ..
HTS on the other hand I have offered to fly/bus multiple times to me. <3
I'm easy (only for her though)
>her -
2017-06-30 at 6:38 AM UTC
-
2017-06-30 at 6:58 AM UTC
-
2017-06-30 at 7:02 AM UTC
-
2017-06-30 at 9:09 AM UTC
-
2017-06-30 at 9:14 AM UTC
Originally posted by inb4l0pht If he's still living in the same area of BC, I'm less than a day's drive away, with no borders to cross. Shouldn't be too hard to make it look like his pretend knowledge of chemistry failed him and resulted in a meth lab explosion.
You're a good poster. You should post more. -
2017-06-30 at 9:36 AM UTC
Originally posted by Captain Falcon http://scikit-learn.org/stable/
Get on it Lenny
freeform text is very high dimensionality data, at least under the obvious model (constituent words, structured or not), traditional classification strategies usually don't work. I mean it's certainly not impossible, it's not trivial. Also because of the significance of a permaban tolerance for false positives is very low. If you want to put together a proof of concept though, go for it. -
2017-06-30 at 10:09 AM UTC
Originally posted by Lanny freeform text is very high dimensionality data, at least under the obvious model (constituent words, structured or not), traditional classification strategies usually don't work. I mean it's certainly not impossible, it's not trivial. Also because of the significance of a permaban tolerance for false positives is very low. If you want to put together a proof of concept though, go for it.
Aka go fuck yourself ain't nobody got time fo dat -
2017-06-30 at 1:40 PM UTC
Originally posted by Lanny freeform text is very high dimensionality data, at least under the obvious model (constituent words, structured or not), traditional classification strategies usually don't work. I mean it's certainly not impossible, it's not trivial. Also because of the significance of a permaban tolerance for false positives is very low. If you want to put together a proof of concept though, go for it.
You could also just implement the suggestion I PMed you. It's a little Orwellian but it is proven to work. Lmk if you want me to send it again. -
2017-06-30 at 1:45 PM UTC