Whenever you're ready, walk me through how you'd approach this.
What loss function should be used in a linear regression model when the data contains many outliers?