I just realized that I missed another category of curriculum learning methods, so I added it in. Now the overview figure looks like this:pic.twitter.com/xofFlKMZfO
Prikaži ovu nit
Humans learn from curriculum since birth. We can learn complicated math problems because we have accumulated enough prior knowledge. This could be true for training a ML/RL model as well. Let see how curriculum can help an RL agent learn:https://lilianweng.github.io/lil-log/2020/01/29/curriculum-for-reinforcement-learning.html …
-
-
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
-
-
Nice post! The area of curriculum learning I’m most excited about is in the generation of entire levels (not just goals) in an open-ended way without explicit goals in mind. i.e. https://twitter.com/hardmaru/status/1150678466481852421?s=21 …https://twitter.com/hardmaru/status/1150678466481852421 …
POET: open-ended coevolution of environments and their optimized solutions@GECCO2019
https://dl.acm.org/citation.cfm?id=3321707.3321799 … pic.twitter.com/Pkei4MoY8e -
Another cool paper about PCG and learning: https://twitter.com/togelius/status/1012726654261702658?s=21 …https://twitter.com/togelius/status/1012726654261702658 …
- Još 2 druga odgovora
Novi razgovor -
-
-
@_Boris_Belousov@Jan_R_Peters We've recently presented an algorithm@CoRl2019, Self-Paced Contextual RL, in which an agent automatically generates a curriculum to learn a specific task or a distribution of tasks and applied it to a real robot, https://arxiv.org/abs/1910.02826 pic.twitter.com/N58ZYVhFBYHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
-
-
We are able to do complex math problems because there is a stage in brain development which develops certain areas in the brain
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
-
-
I know what I'm reading over the weekend
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
-
-
-
Yours is the best blog for modern AI/ML
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.