Exponential smoothing is a weighted$-$moving$-$average forecasting technique that uses a weight, $a,$ such that $0a1.$

When forecasters use exponential smoothing, the new forecast in period $t$ is calculated using the moving average from last period's $(t-1)$ forecast and a percentage of the error from the last period's forecast.

New forecast $=$ Last period's forecast $+$ weight $($Last period's demand $-$ Last period's forecast$)$

$F_t=F_{t-1}+a(A_{t-1}-F_{t-1})$

A backpack manufacturer needs to forecast demand in week $5.$ The backpack demand is shown in Table $3.$