Reinforcement Learning for Execution under Dynamic Fees in a Closed-Loop DEX Simulator

By Wen-Ting Wang

Published 2026-07-12

Everscope rating

1710.6

Relevance to quantitative trading

8 / 10

Implementation complexity

7 / 10

Reproducibility

5 / 5

About this paper

Methodology: Deep Q-Network (DQN) for Optimal Execution in a Closed-Loop DEX Simulator. Problem types: Reinforcement Learning, Algorithmic Execution, Optimization, Market Making, Portfolio Optimization.

arXiv:2607.10960 · Code · Paper rankings

Open the interactive Everscope explorer for full analysis, charts, and paper battles.