Outlier detection in regression: conic quadratic formulations

Andrés Gómez, José Neto

In many applications, when building linear regression models, it is important to account for the presence of outliers, i.e., corrupted input data points. Such problems can be formu…