I want to forecaste the call number everyday for a call-center. Now I
have removed the influence of the fluctuation with some method, so
only thing left is to analyze the trend of the call number every day.
I have thought of two ways: regression and HoltWinters smooth. But
when I use regression, I find some day's call number will bcome
negative, which is obviously unreasonabe. If I use HoltWinters
smooth ,I let the gamma parameter equal to 0 because I needn't analyze
the seasonal fact, but the result shows that the increasing trend is
too big. I adjust the parameter alpha and beta ,which makes the result
better,but I don't know if it's right. In this case of situation,which
way is better? Or is there another appropriate method?