Open-source project
Open-source project
FinRL
An open-source framework for experimenting with deep reinforcement learning agents in financial markets.
- Provider
- AI4Finance Foundation
- Access
- Open source
- Primary interface
- Deep reinforcement learning framework
AI4Finance Foundation
API / Skill
Research
Prototype reinforcement-learning policies
Research results are highly sensitive to reward design, costs and train-test separation.
Open source · Deep reinforcement learning framework
Best used for
Prototype reinforcement-learning policies
Compare agents in market environments
Study portfolio allocation research
Quick start
git clone https://github.com/AI4Finance-Foundation/FinRL.git
cd FinRL
pip install -e .
# Start from the maintained examples and validate
# every result on a strictly held-out period.A practical workflow
- 1Confirm the source and license
- 2Run the smallest useful request
- 3Validate fields and missing values
- 4Cache responsibly and document assumptions
Before you build
Research results are highly sensitive to reward design, costs and train-test separation.
Frequently asked questions
What can FinRL do in a real workflow?
Prototype reinforcement-learning policies Compare agents in market environments Study portfolio allocation research
What access and compatibility does FinRL require?
Open source. Deep reinforcement learning framework.
What is the main implementation risk with FinRL?
Research results are highly sensitive to reward design, costs and train-test separation.