As the original author of what became "BFGS" in optim(), I would point out that
BFGS is a catch-all
phrase that should be applied only to the formula used to update EITHER (my
emphasis) the approximate
Hessian or the INVERSE approximate Hessian. The starting approximation can vary
as well, alon
Rob J Hyndman gives great explanation here
(https://robjhyndman.com/hyndsight/estimation/) for reasons why results
from R's arima may differ from other softwares.
@iacobus, to cite one, 'Major discrepancies between R and Stata for
ARIMA'
(https://stackoverflow.com/questions/22443395/major-dis
2 matches
Mail list logo
