Benchmarking Robustness of Deep Reinforcement Learning approaches to Online Portfolio Management
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2023
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866909072532439040 |
|---|---|
| author | Velay, Marc Doan, Bich-Liên Rimmel, Arpad Popineau, Fabrice Daniel, Fabrice |
| author_facet | Velay, Marc Doan, Bich-Liên Rimmel, Arpad Popineau, Fabrice Daniel, Fabrice |
| contents | Deep Reinforcement Learning approaches to Online Portfolio Selection have grown in popularity in recent years. The sensitive nature of training Reinforcement Learning agents implies a need for extensive efforts in market representation, behavior objectives, and training processes, which have often been lacking in previous works. We propose a training and evaluation process to assess the performance of classical DRL algorithms for portfolio management. We found that most Deep Reinforcement Learning algorithms were not robust, with strategies generalizing poorly and degrading quickly during backtesting. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2306_10950 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | Benchmarking Robustness of Deep Reinforcement Learning approaches to Online Portfolio Management Velay, Marc Doan, Bich-Liên Rimmel, Arpad Popineau, Fabrice Daniel, Fabrice Machine Learning Portfolio Management Deep Reinforcement Learning approaches to Online Portfolio Selection have grown in popularity in recent years. The sensitive nature of training Reinforcement Learning agents implies a need for extensive efforts in market representation, behavior objectives, and training processes, which have often been lacking in previous works. We propose a training and evaluation process to assess the performance of classical DRL algorithms for portfolio management. We found that most Deep Reinforcement Learning algorithms were not robust, with strategies generalizing poorly and degrading quickly during backtesting. |
| title | Benchmarking Robustness of Deep Reinforcement Learning approaches to Online Portfolio Management |
| topic | Machine Learning Portfolio Management |
| url | https://arxiv.org/abs/2306.10950 |