Mean Absolute Error (MAE) quantifies how far anticipated qualities are away from watched esteems.
It’s somewhat unique in relation to Root Mean Square Error (RMSE). overall, it’s just a few basic steps and applying the formula in Excel. MAE plays out the accompanying 2 computations:
MAE sums the absolute value of the residual
Divides by the number of observations.
As portrayed above, here is the MAE Formula:
How about we go over an example of how to compute MAE in Excel. So as to finish this instructional exercise, you will require a lot of watched and anticipated qualities. Additionally, we accept you have Microsoft Excel.
1 Enter headers in the main column of Excel
In A1, type “observed value”. In B2, type “predicted value”. In C3, type “difference”. These are just headers to help identify which values belong to predicted or observed.
2 Place values in columns
If you have 10 observations, place these observed values in cells A2 to A11. In addition, you will type in predicted values from B2 to B11. But you can enter as many values as you’d like in these columns and adjust the following steps accordingly.
3 Find the difference between observed and predicted values
In column C2 to C11, subtract observed value and predicted value. C2 will use this formula:
You will have to copy and paste this formula all the way down to the last row.
4 Calculate the mean absolute error (MAE)
In cell D2, we can calculate MAE by using the formula below:
After entering this code in Excel, cell D2 is the Mean Absolute Error value.
How to use MAE in GIS?
MAE measures the contrast between determined and watched values. For instance, you could analyze satellite-inferred soil dampness esteems and contrast them with what was collected in the field.
For this situation, the satellite-inferred soil moisture values are the forecasted values. At long last, the system of stations on the ground estimating the genuine soil dampness esteems are observed values.