I'm working with h2o in RStudio, specifically the h2o.glm() model, and I'm trying to manually calculate the std_error of my coefficients. Does anyone know what formula is used?
I tried using the formula SE(Beta Hat) = sqrt(MSE/sum((Xi-Xbar)^2)) with the code below. I don't expect the answer to be the exact SE since it's an estimation, but the answer I got wasn't close.
teststack <- model_stack$variable
denom <- sum((teststack$variable-mean(teststack$variable))^2)
se <- sqrt(h2o.mse(model_fit_uc)/denom)
se
