So I'm trying to build a Reinforcement Learning based Agent to do Portfolio allocation on time-series data using an ensemble of algorithms.
What I don't understand is how do I go about creating this project. There are so many things to take care of for instance:
- taking input data
- creating the ensemble using high-quality implementations of DRL algorithms
- using the ensemble to take decisions and so on To be honest, I don't even have absolute clarity about my project and initially thought I will get a better idea as I get started. Do people build a roadmap of the things to want to achieve? Having not created any project (of this magnitude) from scratch before, the overall complexity seems intimidating. Any tips are welcome :)
Top comments (0)