ASSIGNMENT $2$

A company has been using a single exponential smoothing model with an alpha of $0\mathrm{.}2$ to

forecast weekly lorry sales. Given the data below, would a naive forecast have provided

greater accuracy? Explain your answer by calculating MAD, MSE and MAPE. Assume

that there is an initial exponential forecast of $60$ units in period $2($i$.$e$.,$ no forecast for

period $1)$; between MAD, MSE, and MAPE, which measure of forecasting error would

you prefer to use and why?

A manager uses the given linear regression model $y(t)$ to predict demand: $Y_t=20+4t.$

Over the past $8$ periods, demand has been as outlined in the table below. Are the results

acceptable? Explain your answer after calculating control limits for forecasting errors.

Please use control limits covering $95\%$ of the distribution of forecasting errors. Assume

that forecasting errors are normally distributed.