Analyse past data of real-estate demand and use time series forecasting to predict future demand After cleaning an existing data-set containing information about various real-estate micro-markets in Chennai, machine learning algorithms had to applied on them to investigate the factors affecting demand. When these factors were diagnosed, using ARIMA and VARMAX models time series forecasting was performed on them to evaluate future values of demand, if the accuracy was over 80% that model was deemed fit for further determination and thus future values of demand were calculated/