Kaixhin/Atari

Implement Pop-Art

Open

#6 geöffnet am 13. Apr. 2016

Auf GitHub ansehen
 (0 Kommentare) (0 Reaktionen) (0 zugewiesene Personen)Lua (72 Forks)github user discovery
enhancementhelp wanted

Repository-Metriken

Stars
 (263 Stars)
PR-Merge-Metriken
 (Keine gemergten PRs in 30 T)

Beschreibung

Learning functions across many orders of magnitudes introduces Preserving Outputs Precisely, while Adaptively Rescaling Targets (Pop-Art). In summary it normalises outputs across orders of magnitudes and gets rid of the clipping (i.e. counting) rewards heuristic for Atari games. The normalisation is also better for non-stationary problems, i.e., any decent real world problem.

The below is a picture of extra notes from the authors, next to their poster at NIPS 2016: img_20161207_190428

Contributor Guide