This is the official implementation of the KDD 2024 "MacroHFT: Memory Augmented Context-aware Reinforcement Learning On High Frequency Trading". https://arxiv.org/abs/2406.14537
To run the demo code:
You may first download the dataset from Google Drive:
https://drive.google.com/drive/folders/1AYHy-wUV0IwPoA7E1zvMRPL3wK0tPNiY?usp=drive_link
and put the folder under data folder.
Run scripts/decomposition.sh for data decomposition and labeling.
Run scripts/low_level.sh for low-level policy optimization.
Run scripts/high_level.sh for meta-policy optimization.