Q5 25 Marks a Briefly explain the difference between time se

Q5. [25 Marks] a. Briefly explain the difference between time series techniques and causal forecasting method. Provide example to support your answer (8 Marks) b. A 24-hour coffee/donut shop makes donuts every eight hours. The manager must forecast donut demand so that the bakers have the fresh ingredients they need. Table 4 shows the actual number of glazed donuts (in dozens) sold during 3-6 of June Table 4 Actual number of glazed donuts (in dozens) sold during 3-6 of June Shift Demand(dozens) Date Shift Demanddozens)Date June 3Da 59 47 35 64 43 39 June 5 62 46 42 64 50 40 Evenin Evening 1t June 4Da June 6 Evenin Evening- i. Forecast the demand for glazed donuts for the three shifts on June 7.(12 Marks) ii. Make comments on the result:s (5 Marks)

Solution

A.

B.

So we introduce dummy variables to get the seasona indices:

Now we find the regression:

For demand as Y and time series and S1,S2 and S3 as X fr regression.

The result is:

So we see there is error, so we romove s3 from the regression X

And run again the regression analysis.

So the Equation for forecast is:

Forecast = 35.42+ 0,478 t+ 24.206*S1+ 7.978 * S2

So now using the equation we get

the forecast:

ii.

.

So, we see the MAD is not very high as there is clear seasonality, for Day, evening and Night.

But the factor for S3 is ignored and we see from the below P value that, Intercept and S1 is also less relevant.

	Time series	Causal forecasting
	This method takes the seasonality and cyclic trend to consideration	This takes the averaged values of observations over a period
	The methods are slighty difficult to implement and the cyclicity needs to be determined	The methods are generally easy and computation simple
	Time period of forecast is generally short	It is used for medium term forecasts
Example
	Sales of cold drinks in summer, fall etc	Sale of computers, which is not influenced by seasonality