New Working Paper: Missing Data in Real Estate

• Leow, K. and T. Lindenthal. “Machine Learning and Missing Data in Real Estate”
  • NEW WORKING PAPER, FEEDBACK MUCH APPRECIATED. Real estate research tends to be plagued by missing data. We show that prediction accuracy can increase by incorporating observations with missing predictors in the case of commercial real estate. We also show that missing data may not be occurring at random, which makes it more important to incorporate all observations into a prediction model, be it complete or not. Finally, we show that when one incorporates missing data into training sets, prediction outcomes can go into opposite direction