6 Which of the following gives robust tuning parameters. a. Integral of square error
b. Integral of absolute value of error
c. Integral of time weighted absolute error
d. None of these
The distinctive feature of the resulting tuning procedure is the incorporation of the desired robustness level as an explicit design parameter. Although robust divergence, such as density power divergence and -divergence, is helpful for robust statistical inference in the presence of outliers, the tuning parameter that controls the degree of robustness is chosen in a rule-of-thumb, which may lead to an inefficient inference.
Comments
Leave a comment