Title: Spatial bootstrapped microeconometrics: Forecasting for out-of-sample geo-locations
Authors: Katarzyna Kopczewska - University of Warsaw (Poland) [presenting]
Abstract: Spatial econometrics for mass point geo-locations has a limited possibility of forecasting with a calibrated model for a new out-of-sample geo-point. This is because the spatial weights matrix $W$ is defined for in-sample observations only, as well as the computational complexity. The aim is to propose a novel methodology to calibrate both space and model relationships by using bootstrap and tessellation. Bootstrapping enables the calibration of the econometric model without the need for estimation on the whole dataset. Partitioning Around Medoids (PAM) algorithm finds the best points representation in the bootstrapped set of models and generates the medoids coefficients. Tessellated points in the selected best model allow for a representative division of space. New out-of-sample points are assigned to tiles and linked to $W$ as a replacement for original point. The quality of the forecast is tested for the different scenarios of this bootstrap procedure. This efficient procedure supports the big data geo-located point data and makes feasible a usage of calibrated spatial models as a forecasting tool for out-of-sample data. This methodology will find its applications in real estate market forecasting as well as models of business location.