Robust Parametric Approaches
M-estimators do not necessarily relate to a density function and so are not fully parametric. Fully parametric approaches to robust modeling and inference, both Bayesian and likelihood approaches, usually deal with heavy tailed distributions such as Student's t-distribution.
For the t-distribution with degrees of freedom, it can be shown that
.
For, the t-distribution is equivalent to the Cauchy distribution. Notice that the degrees of freedom is sometimes known as the kurtosis parameter. It is the parameter that controls how heavy the tails are. In principle, can be estimated from the data in the same way as any other parameter. In practice, it is common for there to be multiple local maxima when is allowed to vary. As such, it is common to fix at a value around 4 or 6. The figure below displays the -function for 4 different values of .
Read more about this topic: Robust Statistics
Famous quotes containing the words robust and/or approaches:
“It is an immense loss to have all robust and sustaining expletives refined away from one! At ... moments of trial refinement is a feeble reed to lean upon.”
—Alice James (1848–1892)
“I should say that the most prominent scientific men of our country, and perhaps of this age, are either serving the arts and not pure science, or are performing faithful but quite subordinate labors in particular departments. They make no steady and systematic approaches to the central fact.... There is wanting constant and accurate observation with enough of theory to direct and discipline it. But, above all, there is wanting genius.”
—Henry David Thoreau (1817–1862)