Exponential Smoothing

Question 1

Exponential Smoothing Equation

Answer 1

Ft+1 = Ft + α * (Yt - Ft) or Ft+1 = α*Yt + (1-α)Ft

Question 2

True/False The exponential smoothing method weighs recent observations more heavily than previous ones.

Answer 2

True

Question 3

True/False Exponential smoothing requires minimum amount of record keeping for past data

Answer 3

True

Question 4

True/False Exponential smoothing is appropriate where there is a predictable upward or downward trend.

Answer 4

False

Question 5

True/False In exponential smoothing, a higher value of α gives more weight to recent values.

Answer 5

True

Question 6

True/False In exponential smoothing, a lower value of α gives more weight to recent values.

Answer 6

False

Question 7

True/False Exponential smoothing is appropriate where there is no predictable upward or downward trend.

Answer 7

True

Question 8

Should you use single exponential smoothing or double exponential smoothing when there is a trend in the data?

Answer 8

Double Exponential smoothing

Question 9

What is another name for Double Exponential Smoothing?

Answer 9

Holt’s Exponential Smoothing

Question 10

True/False Holt’s Exponential Smoothing is better than Single Exponential Smoothing when there is a trend in the data.

Answer 10

True

Question 11

True/False Single Exponential Smoothing is better than Holt’s Exponential Smoothing when there is a trend in the data.

Answer 11

False

Question 12

What is a drawback of Holt’s Exponential Smoothing?

Answer 12

Forecasts may not be good after a few periods

Question 13

What is β in Holt’s Exponential Smoothing?

Answer 13

A trend factor parameter used to adjust for trend

Question 14

What are the two equations associated with Holt’s Exponential Smoothing?

Answer 14

Smoothing Level Adjustment

Trend Adjustment

Question 15

Holt’s Exponential Smoothing

Smoothing Level Adjustment Equation

Answer 15

St = α(Current Value) + (1-α) (Level + Trend Adjustment)(t-1)

Question 16

Holt’s Exponential Smoothing

Forecast Equation

Answer 16

F(t+m) = St + mTt

Question 17

How is seasonal data identified?

Answer 17

Seasonal data is identified by finding a persistent pattern that occurs at regular time intervals.

Question 18

What is the Holt-Winters Model useful for tracking?

Answer 18

Seasonal Patterns

Question 19

Holt’s Exponential Smoothing

Trend Adjustment

Answer 19

Tt = β (St - S(t-1)) + (1 - β) T(t-1) Tt = β(Change in level in the last period) + (1-β) (Trend Adjustment)(t-1)

Question 20

Holt-Winters Model

Smoothing Level Adjustment Equation

Answer 20

Question 21

Holt-Winters Smoothing

Trend Adjustment

Answer 21

Question 22

Holt-Winters Model

Seasonal Adjustment

Answer 22

Question 23

Holt-Winters Model

Forecast Equation

Answer 23

Question 24

True/False

Exponential Smoothing Method is similar to weighted moving average

Answer 24

True

Question 25

True/False

In Exponential Smoothing one only needs “todays observed” and “yesterdays forecasted value of today” to calculate forecast for tomorrow

Answer 25

True

Question 26

True/False

A higher value of α gives higher weight to the previous forecast and hence a smoother response to the current observation

Answer 26

False

Question 27

True/False

There are two equations associated with Double Exponential Smoothing, one with Smoothing Level and one with Seasonality

Answer 27

False

Question 28

True/False

Double Exponential Smoothing also known as Holt’s Exponential Smoothing it uses a second constant β such that 0 < β < 1 to track trending data better

Answer 28

True