Buad306 quiz 1

The forecast in the previous period was 70. The observed value in the previous period was 90. Using an exponential smoothing model, the forecast in the current period is 73. What value of alpha was used? Carry your calculation to 2 decimal places. If you do not, your answer will be marked incorrect.

.15 (Duplicate of 4a on Group Assignment, with numbers changed)

You use an exponential smoothing model with alpha = 0.4 to forecast staffing needs for the next 4 weeks. For the 4th week forecast, what weight is placed on your initial week 1 forecast F1? Round to exactly 3 decimal places! If not, your answer will be marked incorrect.

0.064 Same problem as 4b on group assignment, with numbers changed. Also, see line 4 on slide 3 of 'what did we learn' and problem 9 on Practice Quiz.

A random-walk forecast (also known as a naive forecast) is just an exponential-smoothing forecast with alpha equal to what value? Write a number - not a word.

1

The scatterplot represents the number of daily calls to a credit card call center during the past 6 weeks (42 data points). The trend line is shown. How many calls can you expect on day 45 ?

278 (just plug x into the given equation)

Your forecast has underestimated for the first 20 periods. Your calculate your forecast errors as (Forecast - Actual). Based on those 20 periods, your BIAS value is -11. What is the MAD?

11 BIAS is the average of the negative errors. MAD is the average of the absolute value of each of those negative errors. SO, MAD = - (BIAS)

Suppose you are using an exponential smoothing model to forecast raw material availability. You estimate F1 to be 110. If you use an alpha value of zero, what is the forecast in period 4?

110 Forecast for this period = Alpha (actual in last period ) + (1-alpha)(forecast in last period. If alpha = 1, then forecast in this period is just the forecast in the last period -- which is the forecast in the period before that ... which is just F1. F2 = F1 F3 = F2 F4 = F3 etc So the forecast in any future period is just F1. This was a 'must think about it' problem but it's only worth 3 points.

intercept: 47,783.08396 mpg: -846.0997321 cargo capacity: 198.9772451 What is the forecasted value of an SUV that gets 22 miles-per-gallon and has a cargo capacity of 37 square feet?

36,531

The scatterplot above shows the number of daily calls to Citi Bank's call center for the past 6 weeks. The trend line is shown. The number of calls on Saturday tends to be higher than normal with a seasonal index of 1.56. Day 45 is a Saturday. What is the forecasted number of incoming calls for day 45?

459 Slides 7 - 9 on lecture about trend and seasonality. Also, Q12 on practice quiz.

Use a 3 - point moving average to forecast the number of ER visits for Saturday. mon- 45 tue- 40 wed- 42 thur- 44 fri-52

46 Group Assignment, Problem 7 on Practice Quiz. You're just taking the average of 2 numbers for a 2-pt MA or the average of 3 numbers for a 3-pt MA.

Demand for an item is seasonal. A summer seasonal index of .75 means

Actual sales are 25% below average If you multiply a number by .75, you're left with 75% of the number so you're decreasing its value by 25%.

The forecasted inventory level in week 1 was 100. The actual inventory level in week 1 was 80. Use an exponential smoothing model with α = 0.5 to forecast the inventory level in week 2.

Computing a forecast using exponential smoothing. You have seen this problem multiple times in the examples in class, the Group Assignment and recommended problems from text. Forecast in week 2 = alpha (80) + (1-alpha)(100)

Which of the following would NOT be part of a supply chain?

Customs station at a port

Forecasts based on a weighted average tend to exhibit more variability than the actual data. (T or F)

False ("more" should be "less")

Exponential-smoothing models and moving averages should be used on time series data that exhibit

Only a random component

You've used an exponential smoothing model to forecast for period 4. The forecast in period 4 is a weighted average of

The actual values for periods 1, 2, 3 and the initial period 1 forecast F1 [ discussed on the lecture on exponential smoothing. Also see slide 3 (line 3) on the 'what did we learn' lecture.]

Suppose you use an exponential smoothing model to forecast data that has a consistent upward trend. Which of the following statements is FALSE?

The forecast will always overestimate.

Which of the following would NOT be considered a service industry:

a bakery

A BIAS value very close to zero means that you have an excellent forecast.

false

You computed a forecast for 4 periods and had errors of -3, -10, 10, 11. You computed your errors as (Actual - Forecast). Your forecast ["overestimated", "neither overestimated nor underestimated", "underestimated"] by an average of ["0", "3.5", "4", "5", "2"] units per period.

underestimated 2 Problem 6 on Practice Quiz. Similar problem illustrated on Slides 14 and 16 of Sept 7th lecture.