Publikationen Mitarbeiter
- Achim Schweikard
- Sven Böttger
- Ralf Bruder
- Tolga-Can Callar
- Niclas Erben
- Floris Ernst
- Maria Henke
- Marius Krusen
- Felix von Haxthausen
- Christian Janorschke
- Ngoc Thinh Nguyen
- Jonas Osburg
- Holger Schwegmann
- Daniel Wulff
- Jingyang Xie
- Honghu Xue
Publikationen ehemalige Mitarbeiter
- Mohammed Al-Badri
- Rehman Ansari
- Norbert Binder
- Christoph Bodensteiner
- Alexander Derksen
- Robert Dürichen
- Markus Finke
- Verónica García Vázquez
- Fernando Gasca
- Hamid Hadjar
- Jannis Hagenah
- Maximilian Heinig
- Svenja Ipsen
- Phillip Jauer
- Ivo Kuhlemann
- Jirapong Manit
- Volker Martens
- Lars Matthäus
- Christoph Metzner
- Lukas Ramrath
- Lars Richter
- Nils Rottmann
- Elmar Rückert
- Alexander Schlaefer
- Osama Shahin
- Birgit Stender
- Patrick Stüber
- Tiberiu Viulet
- Benjamin Wagner
- Christian Wilde
- Tobias Wissel
Sample-Efficient Covariance Matrix Adaptation Evolutional Strategy via Simulated Rollouts in Neural Networks
| Type of publication: | Artikel in einem Konferenzbericht |
| Publikationsstatus: | Akzeptiert |
| Jahr: | 2020 |
| Verlag: | Proceedings of International Conference on Advances in Signal Processing and Artificial Intelligence (ASPAI) |
| Ort: | Berlin, Germany |
| URL: | https://cps.unileoben.ac.at/wp... |
| DOI: | https://cps.unileoben.ac.at/wp/ASPAI2020Xue.pdf |
| Abriss: | Gradient-free reinforcement learning algorithms often fail to scale to high dimensions and require a large number of rollouts. In this paper, we propose learning a predictor model that allows simulated rollouts in a rank-based black-box optimizer Covariance Matrix Adaptation Evolutional Strategy (CMA-ES) to achieve higher sample-efficiency. We validated the performance of our new approach on different benchmark functions where our algorithm shows a faster convergence compared to the standard CMA-ES. As a next step, we will evaluate our new algorithm in a robot cup flipping task. |
| Schlagworte: | |
| Autoren: | |
|
Anhänge
|
|
|
|
| [Bib|RIS] | |
|
|
|
