Forecasting Flashcards
Recently, you have been responsible for setting up a call centre that is essential to transform your company to a service-oriented company. To efficently allocate resources, you must now define the target size of the call center. Your company has hardly any experience in such endeavours. However, a certain correlation between the number of key products sold and the number of incoming calls can be assumed.
Which method would you prefer first and foremost to determine the size of the call centre?
Qualitative Method
Causal Method
Time Series Method
causal method
Which forecasting method is initially less suitable when historical data is available?
Which answer is correct or rather applies best:
Quantitative Method
Time Series Method
Causal Method
Qualitative Method
Qualitative Method
Please evaluate the following statements regarding forecasting errors.
Which answers are correct or rather apply best (multiple answers possible):
A) The MAD (Mean Absolute Deviation) does not punish larger deviations more severely than smaller deviations.
B) The simple error has the advantage that positive and negative values cancel each other out. Therefore, it is suitable and useful for quality determination.
C) With the exception of the MSE (Mean Squared Error), all forecast errors are strongly dependent on the order of magnitude.
D) A MAPE (Mean Absolute Percentage Error) of 18 percent means that on average the estimate is 18 percent off from the measured value.
A;D
Which statements are correct? (multiple answers possible)
A) The Mean Absolute Percentage Error (MAPE) depicts how many percent the estimated value deviates from the actual value on average.
B) The Mean Squared Error (MSE) punishes especially large deviations.
C) The MAPE can also be negative.
D) The Mean Absolute Deviation (MAD) can also be negative.
E) The MSE cannot be negative by definition.
F) If the Average Error is 0, the MSE is always also 0.
A;B;E
Determine the quality of the following forecast: 1.1.d_Measuring_Forecast_Error_A_Exercise.xlsxPreview the document
Separate the decimal places from the whole number by a dot.
Squared Error in July:
Mean Absolute Percent Error (MAPE):
Answer 1:
81
Answer 2:
23.3%
Based on past consumption data from Wallace Garden Supply, determine the forecast for January of the following year. Use a Moving Average over 4 periods.
You will find the necessary information in an Excel sheet under the following link: 1.2.b_Moving_Average_A_Exercise.xlsxPreview the document
Forecast for January in the following year:
16,17,18,19,20
MAPE:
11,1%-11,5%
Answer 1:
18
Answer 2:
11.3%
What happens if the number of periods used to calculate a Moving Average is decreased?
Which answer is correct or rather applies best:
A) The tendency to overestimate generally increases
B) The tendency to underestimate generally increases
C) With a rising trend, the tendency to underestimate decreases
D) The forecast reacts more slowly to fluctuations in the measured data
E) The forecast gets worse
C
In the file 1.2.c_Weigthed_Moving_Average_A_Exercise.xlsxPreview the document you will find the monthly sales data of a shampoo over three years. Based on this data, calculate the demand forecast for January of the fourth year. Use a Weighted Moving Average for this purpose over 5 periods.
Note: Please use the Solver, (1) leave the existing values as starting weights, (2) optimize the MAPE and (3) assure that the sum of weigths equals 1.
You may not see the Solver tool in Excel and you may need to activate it first. In Excel, click on “File” in the upper left corner, then on “Options” and then on the tab “Add-Ins”. Activate the Solver by clicking on “Go”, checking the Solver tool and then clicking on “Ok”.
Forecast for January in year 4:
669
In the file 1.2.d_Exponential_Smooting_A_Exercise.xlsxPreview the document (Links to an external site.) you will find the monthly sales data of shampoo over three years. Based on this data, calculate the demand forecast for January of the fourth year. Use the Exponential Smoothing Method and build upon the Excel Solver (optimize MAPE).
Forecast for January in year 4:
652
In the file 1.3.a_Linear_Trend_Analysis_A_Exercise.xlsxPreview the document you will find the weekly demand for a spare part for a total of 60 weeks. Calculate the demand forecast for week 1 based on this data using Linear Trend Analysis.
Forecast for week 61:
650
What is the objective of a linear trend analysis?
Which statements are correct? (multiple answers possible)
A) The representation of linear data using three constants
B) Minimize the distances between the data points and the linear line (sum of squared errors)
C) Maximize the distances between the data points and the linear line (sum of squared errors)
D) Find out if data show a linear trend
E) The representation of linear data using two constants, of which one is the time
B;D
Please evaluate the following statements on Seasonality Analysis.
Which answers are correct or rather apply best (multiple answers possible):
A) The seasonal ratio is the average of all seasonal indices for a particular month or quarter.
B) The seasonal index of 0.91 indicates that the respective month or quarter shows on average over all years 9% less demand than the average demand.
C) The seasonal ratio of 0.64 indicates that the respective month or quarter shows on average over all years 36% less demand than the average demand.
D) The seasonal index shows the average demand of a season relative to the total average demand.
B;D
In the file 1.3.b_Seasonality_Analysis_A_Excercise.xlsxPreview the document you will find the monthly sales figures of wine (in 1000 litres) over 14 years and 8 months. Based on this data, calculate the Seasonal Ratio and the Seasonal Index for the specified periods.
Seasonal Index in January:
0,87
Suppose we calculate the Centered Moving Average (CMA) with k=3 months over a period of 12 months.
Which months can the CMA be calculated for?
Which answer is correct or rather applies best:
1 to 3 3 to 10 2 to 11 1 to 12 10 to 12
2 to 11
Which of the following statements are correct? (multiple answers possible)
A) In contrast to the simple average over all periods, the Centered Moving Average takes account of the trend.
B) The Centered Moving Average also accounts for local trends, while a linear model builds upon an “absolute” trend over the entire period.
C) The Centered Moving Average is, like the linear trend, a straight line with a constant gradient from beginning to end.
D) The Centered Moving Average (of a period t) contains values from the past (< t) and the future (> t).
A;B;D
