Measuring forecast accuracy

1
Measuring forecast accuracy

• What is the accuracy of a particular forecast?
• How to measure the suitability of a particular
  forecasting method for a given data set?

2
Definition of the forecast error

• Error (e) of a forecast is measured as a
  difference between the actual (A) and forecasted
  values (F), that is,
• eA-F,
• or, in a relative form e100 (A-F)/A.
• The error can be determined only when actual
  (future) data are available.

3
Standard statistical measures to estimate errors
(1)

• To preliminary evaluate a forecast and
  suitability of a method, various statistical
  measures may be used. In evaluating forecasts
  obtained by means of the moving average method,
  the following measures may be used

• Mean (average) error (ME)
• Mean absolute error (MAE)
• Mean squared error (MSE)

4
Standard statistical measures to estimate errors
(2 - relative)

• Mean percentage error (MPE)
• Mean absolute percentage error (MAPE)

5
Statistical measures of goodness of fit
In trend analysis the following measures will be
used

• The Correlation Coefficient
• The Determination Coefficient

6
The Correlation Coefficient

• The correlation coefficient, R, measure the
  strength and direction of linear relationships
  between two variables. It has a value between 1
  and 1
• A correlation near zero indicates little linear
  relationship, and a correlation near one
  indicates a strong linear relationship between
  the two variables

7
The Coefficient of Determination

• The coefficient of determination, R2, measures
  the percentage of variaion in the dependent
  variable that is explained by the regression or
  trend line. It has a value between zero and one,
  with a high value indicating a good fit.

8
End