A similar thing is using least squares for linear regression rather than minimizing MAD. In past the argument was that the least squares sum has a closed expression, but with computers even that advantage eliminated.
The nice thing about minimizing MAD is that in typical settings the liner regression line-plane-hyperplane goes through measurement points. As such there is no interpolation and outliers are nicely cut off making the result very robust to measurements errors.
The nice thing about minimizing MAD is that in typical settings the liner regression line-plane-hyperplane goes through measurement points. As such there is no interpolation and outliers are nicely cut off making the result very robust to measurements errors.