# Meta Learning Shared Hierarchies | Two Minute Papers #210

## Метаданные

- **Канал:** Two Minute Papers
- **YouTube:** https://www.youtube.com/watch?v=M_eaS7X-mIw
- **Дата:** 29.11.2017
- **Длительность:** 3:24
- **Просмотры:** 19,609

## Описание

The paper "Meta Learning Shared Hierarchies" and its source code is available here:
https://arxiv.org/abs/1710.09767
https://github.com/openai/mlsh

A video from Robert Miles: https://www.youtube.com/watch?v=MUVbqQ3STFA

We have been experimenting with opening a bitcoin wallet. Let us know if it's working properly and thank you very much for your support! 
Bitcoin: 13hhmJnLEzwXgmgJN7RB6bWVdT7WkrFAHh

We would like to thank our generous Patreon supporters who make Two Minute Papers possible:
Andrew Melnychuk, Brian Gilman, Christian Ahlin, Christoph Jadanowski, Dave Rushton-Smith, Dennis Abts, Eric Haddad, Esa Turkulainen, Evan Breznyik, Kaben Gabriel Nanlohy, Malek Cellier, Marten Rauschenberg, Michael Albrecht, Michael Jensen, Michael Orenstein, Raul Araújo da Silva, Robin Graham, Steef, Steve Messina, Sunil Kim, Torsten Reil.
https://www.patreon.com/TwoMinutePapers

Music: Antarctica by Audionautix is licensed under a Creative Commons Attribution license (https://creativecommons.org/licenses/by/4.0/)
Artist: http://audionautix.com/ 

Thumbnail background image credit: https://pixabay.com/photo-1804496/
Splash screen/thumbnail design: Felícia Fehér - http://felicia.hu

Károly Zsolnai-Fehér's links:
Facebook: https://www.facebook.com/TwoMinutePapers/
Twitter: https://twitter.com/karoly_zsolnai
Web: https://cg.tuwien.ac.at/~zsolnai/

## Содержание

### [0:00](https://www.youtube.com/watch?v=M_eaS7X-mIw) Segment 1 (00:00 - 03:00)

de fellow scholars dit is toen minuut pepers met cadeaus journée fire reinforcement learning is een techniek waar we hebben virtual creature that tries to learn een optimaal zelf actions to maximize er een woord in de changing environment play video games helicopter control and evil optima zing light red sport simulatie een arme man de more awesome ik sample use cases voor de brief moet trainen reinforcement learning from scratch lizzie dat dit type collistar zout hoe de bloed voor search in de space of de simplest noise level actions this not only too crazy begeven yuliyan er is also highly ineffectief requires way more experienced and she'll mijn stoel en die op t knowledge kanaal review for similar tests it can learn de game it was trained and often i wanna superhuman level geef me niet dit te functioneren nu en waren mijn ouders previous knowledge has to be thrown away and this item is wij mars light house us learn it break staan de beek en complex tess kind of sequences of smaller actions bizar code sap policies and can be shared between cask learning to work and krol excellent ik samples of tijd en will likely be used for a variety of different problems and we look to grab it learning kon 0 aanzien tests even if they were significantly from the previous lizzie een probleem not only that wat de search vs offers a policy is can easily be a hundred om marcos smaller than the original search space of all car selection therefore this kind of search is way more efficient than previous techniek of course rating judith selectie of sap policies is challenge ik pik as they have to be robust enough to be helpful and many possible cached het not to specifiek toe want probleem otherwise the loose die utility of you happy social dimension een relatief techniek bij de neem nu al cached programming and it seems that this one is capable of generation net olie over die vriend variations of dus een task breuk rosbief en tasks as well designs were trained to reverse several different measures one after another and quickly via live dat de basic movement directions to be retained creating more general learning a great and is one of the holy grail problems of care research and this one seems to be a proper trapper star tours die feeding het who you're not there yet but it's hard not to be optimist dit moet this incredible weet of progress ietsje wie lake side to see how this er ja impuls over the next few months de source code of this project is also available oh andy foger make sure to check out the channel robert miles to make excellent video zoals een jaar en die wekker mens starting with one of video's de tumor objecten vlieg guaranteed to enjoy the view is to find out why do zie de link in de video description or just click de kat picture papier een keer aan de screen en de moment viel in the danger that make sure to subscribe to this channel deksel wat chinian folio generous support naar see you next time

---
*Источник: https://ekstraktznaniy.ru/video/14549*