GLM Distribution and Link Function


I am trying to predict claim severity of an individual using GLM in R Studio.

I have data of 1 Million individuals, and their claims against them. 95% of the individuals have not recorded any claims, so there are a lot of 0s in the data, and the target variable is also right skewed.

Which distribution/link function would be recommended?

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.