BestNumber5/dqn-monetary-policy-korea: Data and Code for "Integrating Financial Stability Objectives into Monetary Policy using Deep Reinforcement Learning"

Description This release contains the dataset and source code for the manuscript: "Integrating Financial Stability Objectives into Monetary Policy using Deep Reinforcement Learning: An Exploratory Analysis of the South Korean Macroeconomy". Contents: Dataset: Monthly Korean macroeconomic time-series data (Jan 2012 - Mar 2025), including apartment price index, CPI, employment data, household loans, M2, and base rate. Code: VAR model estimation and simulation. Deep Q-Network (DQN) reinforcement learning environment for dynamic Taylor Rule parameter optimization. Scripts for evaluating policy performance across economic scenarios (Stable, Crisis, Inflation, Tightening).