• 0

RMSE

Viewed 830 Times 0 Comments

Root Mean Squared Error.  A method to quantify the positional accuracy of spatial data that measures the difference between observed and predicted values.  When georeferencing imagery, RMSE is calculated with respect to ground control points.  Mathematically, it is the square root of the average of the square of all the error.  In practice, an RMSE of 0 is the target of no error, values less than 1 are desirable.  Since the calculation is based on the squared value, the output error value can inflate quickly as error increases.  Keeping the error low is important for the validity of predictions or models created from the data.

Was this answer helpful ? Yes (0) / No (0)

Leave a Reply