MAPE is defined as which of the following?

MAPE is defined as which of the following?

• A. It is the average of the absolute percentage errors between forecast and actual
• B. It is the square root of the average absolute percentage error
• C. It is the sum of absolute percentage errors
• D. It is the maximum absolute percentage error

Answer: (A)

MAPE measures forecast accuracy by averaging the size of the percentage errors. Each error is the absolute value of (Actual minus Forecast) divided by the Actual value, and then these percentages are averaged (often expressed as a percent). Using the absolute value prevents positive and negative errors from canceling each other, so you’re looking at the typical size of the error relative to what actually happened. This makes the metric easy to interpret and comparable across different scales. The other options don’t fit this concept: taking the square root of the average absolute percentage error isn’t the standard accuracy measure, and would change the scale and interpretation. Summing the absolute percentage errors would depend on how many observations you have, not the average error per item. Focusing on the maximum absolute percentage error looks at the worst-case case rather than typical performance.

MAPE measures forecast accuracy by averaging the size of the percentage errors. Each error is the absolute value of (Actual minus Forecast) divided by the Actual value, and then these percentages are averaged (often expressed as a percent). Using the absolute value prevents positive and negative errors from canceling each other, so you’re looking at the typical size of the error relative to what actually happened. This makes the metric easy to interpret and comparable across different scales.

The other options don’t fit this concept: taking the square root of the average absolute percentage error isn’t the standard accuracy measure, and would change the scale and interpretation. Summing the absolute percentage errors would depend on how many observations you have, not the average error per item. Focusing on the maximum absolute percentage error looks at the worst-case case rather than typical performance.