GMRF based imputation in Average emergency visits

Lets look at the California dataset of average emergency visits. Can we throw a GMRF+SVR regression model at it and impute missing values?
If so, how good are the results?