Incorporating Asymmetric Loss for Real Estate Prediction with Area-level Spatial Data

TL;DR

This paper explores the use of asymmetric loss functions, specifically LINEX and power divergence, in spatial area-level predictions for real estate valuation, highlighting the impact of loss function choice on prediction accuracy.

Contribution

It introduces a methodology for incorporating asymmetric loss functions into CAR models for spatial data, emphasizing parameter selection's importance.

Findings

Asymmetric loss functions significantly affect prediction outcomes.

Parameter choice in asymmetric losses is crucial for optimal spatial predictions.

The methodology improves real estate valuation accuracy.

Abstract

We investigate two asymmetric loss functions, namely LINEX loss and power divergence loss for optimal spatial prediction with area-level data. With our motivation arising from the real estate industry, namely in real estate valuation, we use the Zillow Home Value Index (ZHVI) for county-level values to show the change in prediction when the loss is different (asymmetric) from a traditional squared error loss (symmetric) function. Additionally, we discuss the importance of choosing the asymmetry parameter, and propose a solution to this choice for a general asymmetric loss function. Since the focus is on area-level data predictions, we propose the methodology in the context of conditionally autoregressive (CAR) models. We conclude that choice of the loss functions for spatial area-level predictions can play a crucial role, and is heavily driven by the choice of parameters in the respective loss.