Nonparametric prediction with spatial data
share
We describe a nonparametric prediction algorithm for spatial data. The algorithm is based on a flexible exponential representation of the model characterized via the spectral density function. We provide theoretical results demonstrating that our predictors have desired asymptotic properties. Finite sample performance is assessed in a Monte Carlo study that also compares our algorithm to a rival nonparametric method based on the infinite AR representation of the dynamics of the data. We apply our method to a real data set in an empirical example that predicts house prices in Los Angeles.
READ FULL TEXT