In this chapter, new regression techniques that are particularly suited to the treatment of specific data types are introduced. Starting with the problem of the presence of outliers in our data, we learned to identify outliers and remove them, thus verifying the performance improvements obtained from the template. We have seen that this step can be simply bypassed using a robust regression model.
To reduce outlier effects, we can fit our data using robust least squares regression. Robust regression methods provide an alternative to least squares regression; they attempt to reduce the influence of outlying cases so as to render a better fit to the majority of the data. Robust regression downweighs the influence of outliers, and makes their residuals larger and easier to identify.
Then we explored the Bayesian regression technique. In the Bayesian approach, we see that the interpretation differs substantially, in fact the β coefficients are treated as random variables, rather than fixed...