AI4Finance-Foundation / AI4Finance-Foundation/FinRL
Amend reward function?
Offen
discussion
- Vorherrschende Sprache
- Jupyter Notebook
- Sterne
- 16.3k
- Forks
- 3.5k
- PR-Merge-Kennzahlen
- Keine gemergten PRs in 30 T.
Beschreibung
Currently the tutorial seems to use terminal portfolio value as the reward. I would like to penalize it with some kind of risk measure (such as MDD). Hence my reward would be Terrminal Value - alpha * MDD
I tried to amend self.reward in the Environment for Portfolio Allocation but it seems the resulting portfolio is unchanged when using A2C.
Any idea how I should do it please? Thanks!
Beitragsleitfaden
Für dieses Repository ist kein Beitragsleitfaden indexiert
Bewertung
Dieses Issue wurde noch nicht bewertet.