And yet, this won’t really resolve the issue with long-term projections (50+ years), which is the heart of the issue. An empirical-only method couldn’t possibly predict out 50 years, without itself requiring projections of human behaviour that affect its own “features” per se.
Why not have a @Kaggle competition to predict global temperature? Then we'll see how good those climate models really are.
-
-
-
Though it may be interesting to see how many respectable skeptics would accept the inadequate climate models for 50+ year projections, ahead of naive empirical-only models!
Keskustelun loppu
Uusi keskustelu -
-
-
Data is public so how do you get a hold out?
-
Believe it or not, climate modelers don't use holdouts. They tweak parameters by hand until it gives the results they want. Just what we teach students not to do in ML 101. (And even then their ability to fit past data sucks.)
- Näytä vastaukset
Uusi keskustelu -
-
-
It probably will take 100 years before we know who will win though ...
- Näytä vastaukset
Uusi keskustelu -
-
-
I’ve been suggesting something like that for a while.
- Näytä vastaukset
Uusi keskustelu -
-
On the one hand, this is Pedro Domingos, who says many controversial things which repulse me. On the other hand, it's a good idea! And good ideas deserve to be heard!
Kiitos. Käytämme tätä aikajanasi parantamiseen. KumoaKumoa
-
-
-
Kiitos. Käytämme tätä aikajanasi parantamiseen. KumoaKumoa
-
Lataaminen näyttää kestävän hetken.
Twitter saattaa olla ruuhkautunut tai ongelma on muuten hetkellinen. Yritä uudelleen tai käy Twitterin tilasivulla saadaksesi lisätietoja.