In: Statistics and Probability
How does a forecast based on a weighted moving average differ from that based on exponential smoothing?
weightage moving average is a moving average system with we are giving weightage , like lets consider we are going to find out weighted moving average of a system
first we have to fix up to what previous period we are doing the weighted moving average:
like we fix last n months.
so we give weightage of prior months
latest month we give n weightage, prior to prior we give n-1 like that we give nth prior month the weightage is 1,
then
forecast of the month with weightage moving average= (n*latest month demand +n-1*prior to prior month demand +...............+1* nth prior month demand)/(sum of weightage)
but in exponential smoothening forecasting system we consider all prior data:
then we give one smoothing factor a
and
forecast of nth month with exponential smoothening = a* prior month demand +a*(1-a)* prior to prior month demand + a*(1-a)2 * 3rd prior month demand+.................
main difference between exponential smoothing and weighted moving average is in weighted moving average we consider the limit to how long we have to go back for forecasting but exponential series we consider all demands till last month.