DB $320$ Forecasting Assignment

Actual Sales January $-$ August

Month Sales

January $385$

February $389$

March $394$

April $398$

May $390$

June $410$

July $415$

August $410$

Based on this data, forecast sales for September using:

$1.$ A four month moving average

$2.$ A weighted three month moving average using these weights:

August $0\mathrm{.}6$

July $0\mathrm{.}3$

June $0\mathrm{.}1$

$3.$ Exponential smoothing where $\backslash$alpha $=0\mathrm{.}55($assume the forecast for August was $415)$

$4.$ One of your employees developed two models for forecasting sales and tried to fit the model to past sales

data, as shown below. You're not sure which method is better so you decide to calculate the Mean Absolute

Percentage Error to explain which forecasting method $($Method $1$ or Method $2)$ is "better." Show your work in

calculating the MAPE and make a conclusion as to which is a "better" method for predicting future sales.

Month Mo Counter Actual Sales Method $1$ Method $2$ e$1$ e$2|$e$1||$e$2||$e$1|/$A $|$e$2|/$A

J $1353322340311331130\mathrm{.}0880\mathrm{.}037$

F $2321330345$

M $336233137231-1031100\mathrm{.}0860\mathrm{.}028$

A $43703403643063060\mathrm{.}0810\mathrm{.}016$

M $538236238520-32030\mathrm{.}0520\mathrm{.}008$

J $63773703917-147140\mathrm{.}0190\mathrm{.}037$

J $7392400391-81810\mathrm{.}0200\mathrm{.}003$

A $8404390400$

$4$a$.$ The MAPE $\%$ for Method $1$ and Method $2$: MAPE $\%=$

$4$b$.$ Which forecasting method is better?