1

In socio-economic data, I always found heteroskedasticity that can't be solved using transformation.I had read a paper "Spatial autoregressive models with unknown heteroskedasticity:A comparison of Bayesian and robust GMM approach". In this paper, I know solution the problem of heteroskedasticity in SAR, but I have difficulty to apply Bayesian method for SAR model with heteroscedasticity in R. I have tried using R-Inla, but it turns out that it is for homoscedastic models.

My question is how to implement SAR model with heteroskedasticity using the bayesian approach in R?

kjetil b halvorsen
  • 63,378
  • 26
  • 142
  • 467
Fitri Ramadhini
  • 11
  • 2
  • What is your question? – Richard Hardy Apr 15 '19 at 09:51
  • my question is, how to implement SAR model with heteroscedasticity using the Bayesian approach in r? – Fitri Ramadhini Apr 15 '19 at 14:17
  • How to do something in R is off topic, but you could ask about the general logic of the model which would give you a good enough understanding and allow implementing the model it in any language. – Richard Hardy Apr 15 '19 at 15:22
  • I think all poisson models and perhaps binomial as well are heteroscedastic, right? Anyway, look into GEE models: I've seen an application of GEE models that were used to model poisson overdispersion, and the variance was a function of something – Tomas Jul 21 '19 at 05:45

0 Answers0