CMStatistics 2017: Start Registration
View Submission - CMStatistics
B0186
Title: Assessing the structural risk accounting for ties Authors:  Gianfausto Salvadori - Universita Del Salento (Italy) [presenting]
Fabrizio Durante - University of Salento (Italy)
Roberta Pappada - University of Trieste (Italy)
Abstract: Copulas are useful in quite a few different applications, and especially in environmental sciences, where the variables at play are generally non-independent. Usually, these variables are continuous ones, being times, lengths, weights, and so on. Unfortunately, the corresponding observations may suffer from (instrumental) adjustments and truncations, and may show repeated values (i.e., Ties). As a consequence, on the one hand, a tricky issue of model identifiability may arise, and, on the other hand, the assessment of the risk may be adversely affected. A possible remedy consists of suitable data randomization: three different jittering strategies are outlined. A simulation study is carried out in order to evaluate the effects of the randomization of multivariate observations on the risk assessment when ties are present. In particular, it is investigated whether, how, and to what extent, the randomization may change the estimation of the structural risk by using a coastal engineering example, as archetypical of a broad class of models/problems in engineering applications. Practical warnings and advices about the use of randomization techniques are also given.