Stage-oe-small.jpg

Thema4679: Unterschied zwischen den Versionen

Aus Aifbportal
Wechseln zu:Navigation, Suche
(Die Seite wurde neu angelegt: „{{Abschlussarbeit |Titel=Efficient Deep Reinforcement Learning by Combining Variational Autoencoders with Soft Actor Critic |Vorname=Moritz |Nachname=Nekolla…“)
 
Zeile 7: Zeile 7:
 
|Partner=FZI
 
|Partner=FZI
 
|Forschungsgruppe=Angewandte Technisch-Kognitive Systeme
 
|Forschungsgruppe=Angewandte Technisch-Kognitive Systeme
|Abschlussarbeitsstatus=In Bearbeitung
+
|Abschlussarbeitsstatus=Abgeschlossen
 
|Beginn=2020/08/01
 
|Beginn=2020/08/01
 
|Beschreibung DE=The objective of this work is to develop and evaluate the power of a Variational Autoencoder (VAE) to enforce Reinforcement Learning (RL) on a wide variety of tasks. This should be enabled by automatically extracting important features from the observation and creating a more dense representation. Applying an RL algorithm on this pre-trained latent space could be more likely to converge.
 
|Beschreibung DE=The objective of this work is to develop and evaluate the power of a Variational Autoencoder (VAE) to enforce Reinforcement Learning (RL) on a wide variety of tasks. This should be enabled by automatically extracting important features from the observation and creating a more dense representation. Applying an RL algorithm on this pre-trained latent space could be more likely to converge.
 
}}
 
}}

Version vom 26. Oktober 2020, 11:53 Uhr



Efficient Deep Reinforcement Learning by Combining Variational Autoencoders with Soft Actor Critic


Moritz Nekolla



Informationen zur Arbeit

Abschlussarbeitstyp: Bachelor
Betreuer: Mohammd Karam Daaboul
Forschungsgruppe: Angewandte Technisch-Kognitive Systeme
Partner: FZI
Archivierungsnummer: 4679
Abschlussarbeitsstatus: Abgeschlossen
Beginn: 01. August 2020
Abgabe: unbekannt

Weitere Informationen

The objective of this work is to develop and evaluate the power of a Variational Autoencoder (VAE) to enforce Reinforcement Learning (RL) on a wide variety of tasks. This should be enabled by automatically extracting important features from the observation and creating a more dense representation. Applying an RL algorithm on this pre-trained latent space could be more likely to converge.