here’s a conjecture: biological learning is never unsupervised. Rather, its goal is to learn representations that are useful for future behavior. @AstroKatie @AToliasLab @danilobzdok
-
-
You can break this down into a few possibilities 1) something like this, which is not simply optimizing some parsimonious/simple objective like “compression” or “reconstruction”, and whose goal is downstream task performance, but is still unsupervised: https://arxiv.org/abs/1804.00222
-
yes, exactly that it seems! i should have known jascha already wrote this down :)
Kraj razgovora
Novi razgovor -
-
-
people typically call this self-supervised learning
-
Yes, I've also heard the term "natural supervision". But, at the end of the day, there is no fundamental difference with unsupervised learning algorithms.
Kraj razgovora
Novi razgovor -
-
-
This discussion reminds me of decision-aware model learning in the context of model-based RL. See my papers on Value-Aware Model Learning framework (AISTATS 2017 and NeurIPS 2018). http://sologen.net/papers/VAML%28AISTATS2017%29.pdf … http://sologen.net/papers/IterVAML(NeurIPS2018)(extended).pdf …
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
-
-
Some form of transfer learning could bridge unsupervised and supervised learning components
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
Čini se da učitavanje traje već neko vrijeme.
Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.