Synthetic Data for Article: Alleviating Housing Market Shocks in Real-Time: an Agent-Based Reinforcement Learning Approach

There are two sets of data in this repository, these are: Model Validation Python_Base_100Runs.csv Netlogo_Base_100Runs.csv Python_Ratefall_100Runs.csv Netlogo_Ratefall_100Runs.csv Article Experiment Outputs LTV_80_Ratefall_NORL.csv LTV_100_Ratefall_NORL.csv LTV_80_Ratefall_Shock_Training_RL.csv LTV_100_Ratefall_Shock_Training_RL.csv The model validation datasets compare identical experiments conducted on two different programming frameworks and quantified the similarities between the two model architectures. These results are referenced in the paper in section titled: Quantifying model similarities (validation) The artical experiment outputs compare base case conditions (no reinforcement learning) and reinforcement learning decisions and the housing market reactions to RL agent involvement. Generally, the RL agent adjusts the interest rates to stabilise the housing market indicators. These datasets are purely experimental and will accompany the article mentioned earlier.