Chapter 5 Forecasting
All seasons must be same in terms of number of periods contained. a. True b. False
False
When 'seasonal variations' are considered, the seasons must always go with the four natural seasons. a. True b. False
False
__________ is a measure of accuracy of a forecast model.
MAD
In 4-period-moving average, the forecast of the current period is the average of the sales of ____________________. a. any four past periods b. four most recent past periods c. four earliest past periods
four most recent past periods
A forecast should reflect_______, _______, and _________, It should NOT be misled by _________.
"trend", "seasonal variation", and "business cycle" "random variations (noises)"
Suppose July's forecast was $500 and July's actual sales was $400. What is the absolute forecast error of July? a. $100 b. -$100 c. $500 d. $400
$100
In a regression equation, Y = a + bX, Y is called the________ variable, and X is called _______ variable.
dependent, independent
Suppose July's forecast was $500 and July's actual sales was $400. What is the forecast error of July? a. $100 b. -$100 c. $500 d. $400
-$100
. Suppose there are two business seasons in a year, low season and high season. Let Y=sales of product A; X1=period number (1, 2, 3, ..., 24); X2=1 if the period is in the low season, X2=0 if the period is in the high season. The regression equation derived from the past 24 periods' data is as follows: Y = 3,170 + 18X1 - 350X2. (a) Calculate the forecast of the sales for the next period which is in the low season. (b) Calculate the forecast of the sales for the period two periods from now in future which is in the high season.
.(a)Y=3170+18*25-350*1=3270 26.(b)3170+18*26-350*0=3638
If there are two seasons, high and low, for a product, then _____ dummy variable(s) will be used in the regression model. a. 0 b. 1 c. 2 d. 3
1
For questions 18 - 19: Suppose we have sales data for Jan., Feb., and March as follows: Jan $1,500 Feb $1,000 March $800 18. What is the forecast of April by using 3-month-moving average? 19. What is the forecast of April by using 2-month-weighted-moving-average with weights (3 for most recent past period, 2 for second most recent past period)?
18. 1100 19. 880
If there are two seasons, high and low, for a product, then _____ independent variables will be used in the regression model. a. 0 b. 1 c. 2 d. 3
2
Three business seasons are considered in a state park, hot season (June, July, August), Warm season (March, April, May, September, October) and cold season (November through February). If multiple regression is used to do forecasting of monthly number of visitors based on the past five years monthly data, then there are______ independent variables a. 2 b. 3 c. 4 d. 5 e. 12 f. 60
3
A dummy variable in the regression method for seasonal variations can only take value of 0 or 1. a. True b. False
True
Forecasting is a scientific methodology for predicting the future. a. True b. False
True
Suppose we have the sales data of last 48 months, and we use the 2-month moving average method. To calculate the forecast of the current month, we only need the sales data of last two months. a. True b. False
True
Suppose we have the sales data of last 48 months, and we use the exponential smoothing method with a=0.3 and the initial month's forecast was 580. To calculate the forecast of the current month, we need to calculate the forecast of each of the past 48 months. a. True b. False
True
Given regression equation Y = -350 + 80x, where Y = weekly sales and X series number of weeks (1, 2, ..... 52), is derived from actual sales data in the past year. Calculate the forecast of sales for the 53rd week.
Y = -350 + 80 (53) = $3,890
In a real world application of exponential smoothing, how is the value of a determined? a. It is determined arbitrarily. b. It is given by your boss who has obligation to provide that value. c. You do experiments on various values of a and pick the a that generates smallest MAD.
You do experiments on various values of a and pick the a that generates smallest MAD.
The formula for exponential smoothing is Ft=Ft-1+a(At-1-Ft-1), in which a _________ a. is between 0 and 1 b. is between -1 and 1 c. can be any positive number
is between 0 and 1
When using the exponential smoothing method, if we want our forecasting more responsive to the trend in the data, we should select a _________ a. a. smaller b. larger
larger
When using the n-period-moving average method, if we do not want our forecasting responsive to the random variations or noises in the data, we should select a _________ n. a. smaller b. larger
larger
When using the weighted moving average method, if we think the most recent past period's sales is most relevant for the forecast of the current period, then we should give the most recent past period a _______ weight. a. smaller b. larger
larger
. MAD stands for ____________. a. moving average decision. b. median of absolute deviation. c. mean absolute deviation.
mean absolute deviation
What is the restriction of the values of a and b in regression equation Y = a + bX? a. (a>=0, and b>=0) b. no restriction on a, but b>=0 c no restriction on a or b d (a>=0, and no restriction on b)
no restriction on a or b
'Seasonal variations' refer to _____________. a. random variations without explainable reasons b. recurring variations at certain seasons of a year c. changes due to temperatures
recurring variations at certain seasons of a year
The __________ MAD the better. a. smaller b. larger
smaller
When using the exponential smoothing method, if we do not want our forecasting responsive to the random variations or noises in the data, then we should select a _________ a. a. smaller b. larger
smaller
When using the n-period-moving average method, if we want our forecasting more responsive to the trend in the data, we should select a _________ n. a. smaller b. larger
smaller
Forecast error of March is the difference between ______________. a. the actual sales of March and forecast of March b. the forecast of February and the forecast of March c. the forecast of March and MAD
the actual sales of March and forecast of March
MAD means _____________________. a. the average of absolute forecast errors b. the total of absolute forecast errors c. the average of forecast errors d. the total of forecast errors
the average of absolute forecast errors
In multiple regression equation Y = a + b1X1 + b2X2 + b3X3 + ..... bnXn which is used for doin forecasting with both trend and seasonal variation taken into account,___________ is (are) binary independent variable that can only take value 1 or 0. a. Y b. x1, x2,x3, ..... xn c. Y x1, x2, x3,......xn d. x1 e. x2, x3,....., xn
x2, x3,....., xn
Let Y = sales of product A; X1 = period series number; X2 = 0 if it is not season 2. Season 1 is composed of November and December. season 2 is composed of the other ten months. The regression equation derived fro the past 18 months data is as follows: Y= 2,000 + 10x1 - 40x2 currently we are at the very endow month 18 which is November. a. Calculate the forecast of the sales for the next month which is December b. calculate the forecast of the sales for next January
y = 2000 + 10 (19) - 40 (0) = 2190 y= 2000 + 10 (20) - 40 (1) = 2160
