# Heterogeneous Regression Models for Clusters of Spatial Dependent Data

## TL;DR

This paper introduces a Bayesian clustered regression approach using Dirichlet processes to identify clusters of regions with similar economic characteristics in spatial data, demonstrated through simulations and a housing cost case study.

## Contribution

It presents a novel Bayesian method for detecting covariate effect clusters in spatial data, accommodating an unknown number of clusters.

## Key findings

- Effective in identifying meaningful spatial clusters
- Demonstrated robustness in simulation studies
- Applied successfully to real housing data

## Abstract

In economic development, there are often regions that share similar economic characteristics, and economic models on such regions tend to have similar covariate effects. In this paper, we propose a Bayesian clustered regression for spatially dependent data in order to detect clusters in the covariate effects. Our proposed method is based on the Dirichlet process which provides a probabilistic framework for simultaneous inference of the number of clusters and the clustering configurations. The usage of our method is illustrated both in simulation studies and an application to a housing cost dataset of Georgia.