Reinforcement Learning for Optimal Execution when Liquidity is Time-Varying
By Andrea Macrì, Fabrizio Lillo
Rating
1478
Battle Count: 74
Relevance
9/10
Highly relevant for developing adaptive execution strategies in realistic market conditions with time-varying liquidity
Implementation Complexity
7/10
Requires implementation of DDQL algorithm and market simulation environment, but methodology is well-described
Reproducibility
4/5
Detailed methodology and parameters provided, but no code repository linked
About this paper
Methodology: Double Deep Q-Learning (DDQL). Problem types: Algorithmic Execution, Optimization.
The interactive Everscope explorer (charts, battles, favorites) loads below.