132k views
3 votes
Week 1 2 3 4 5 6

Value 18 14 16 11 17 15

Calculate the measures of forecast error using the naive (most recent value) method and the average of historical data (to 2 decimals). Show your work.

Naive method Historical data
Mean absolute error
Mean squared error
Mean absolute percentage error

Which method provides the most accurate forecasts?

A. Naive method

B. Historical data

User Mmfrgmpds
by
8.6k points

1 Answer

3 votes

Answer:

Explanation:

Let's calculate the measures of forecast error using both the naive (most recent value) method and the average of historical data.

Given time series data:

Week: 1 2 3 4 5 6

Value: 18 14 16 11 17 15

Naive Method:

For the naive method, the forecast for the next period is the most recent value. So, the forecast for Week 7 is 15.

Historical Data (Average) Method:

The forecast for Week 7 using the average of historical data is the average of values in Weeks 1 to 6:

\[ \text{Forecast for Week 7} = \frac{18 + 14 + 16 + 11 + 17 + 15}{6} \]

\[ \text{Forecast for Week 7} = \frac{91}{6} \approx 15.17 \]

Now, let's calculate the measures of forecast error:

Mean Absolute Error (MAE):

\[ MAE = \frac{1}{n} \sum_{i=1}^{n} |Y_i - \hat{Y}_i| \]

Naive Method:

\[ MAE_{\text{Naive}} = |15 - 15| = 0 \]

Historical Data Method:

\[ MAE_{\text{Historical}} = \frac{1}{6} (|18-15| + |14-15| + |16-15| + |11-15| + |17-15| + |15-15|) \]

\[ MAE_{\text{Historical}} = \frac{1}{6} (3 + 1 + 1 + 4 + 2 + 0) = \frac{11}{6} \approx 1.83 \]

Mean Squared Error (MSE):

\[ MSE = \frac{1}{n} \sum_{i=1}^{n} (Y_i - \hat{Y}_i)^2 \]

Naive Method:

\[ MSE_{\text{Naive}} = (15 - 15)^2 = 0 \]

Historical Data Method:

\[ MSE_{\text{Historical}} = \frac{1}{6} ( (18-15)^2 + (14-15)^2 + (16-15)^2 + (11-15)^2 + (17-15)^2 + (15-15)^2 ) \]

\[ MSE_{\text{Historical}} = \frac{1}{6} (9 + 1 + 1 + 16 + 4 + 0) = \frac{31}{6} \approx 5.17 \]

Mean Absolute Percentage Error (MAPE):

\[ MAPE = \frac{1}{n} \sum_{i=1}^{n} \left| \frac{Y_i - \hat{Y}_i}{Y_i} \right| \times 100 \]

Naive Method:

\[ MAPE_{\text{Naive}} = \left| \frac{15 - 15}{15} \right| \times 100 = 0 \]

Historical Data Method:

\[ MAPE_{\text{Historical}} \approx \frac{1}{6} (16.67 + 7.14 + 6.25 + 36.36 + 11.76 + 0) \]

\[ MAPE_{\text{Historical}} \approx \frac{78.18}{6} \approx 13.03 \]

Comparison:

- The naive method has lower errors in MAE, MSE, and MAPE compared to the historical data method.

- Therefore, the naive method provides more accurate forecasts based on the given measures of forecast error.

User Ameer Ali Khan
by
8.4k points