Vorlesen

Research Seminar Series in Statistics and Mathematics

Wirtschaftsuniversität Wien, Departments 4 D4.4.00809:00 - 10:10

Art Vortrag/Diskussion
SpracheEnglish
Vortragende/rClara Grazian (Nuffield Department of Medicine, University of Oxford)
Veranstalter Institut für Statistik und Mathematik
Kontakt katrin.artner@wu.ac.at

Clara Grazian (Nuffield Depart­ment of Medi­cine, Univer­sity of Oxford) about "Baye­sian analysis of semi­pa­ra­metric copula models"

The Insti­tute for Statis­tics and Mathe­ma­tics (Depart­ment of Finance, Accoun­ting and Statis­tics) cordi­ally invites ever­yone inte­rested to attend the talks in our Rese­arch Seminar Series, where inter­na­tio­nally renowned scho­lars from leading univer­si­ties present and discuss their (working) papers.

The list of talks for the winter term 2018/19 is avail­able via the follo­wing link: https://www.wu.ac.at/en/stat­math/resse­minar

Abstract:
Appro­xi­mate Baye­sian compu­ta­tion (ABC) is a recent class of algo­rithms which allows for mana­ging complex models where the likelihood func­tion may be considered intrac­table. Complex models are usually charac­te­rized by a depen­dence struc­ture diffi­cult to model with stan­dard tools. Copula models have been intro­duced as a proba­bi­listic way to describe general multi­va­riate distri­bu­tions by conside­ring the marginal distri­bu­tions and a copula func­tion which captures the depen­dence struc­ture among the compo­n­ents of the vector. While it is often strai­ght­for­ward produ­cing reliable esti­mates of the margi­nals, esti­ma­ting the depen­dence struc­ture is more compli­cated, in parti­cular in high dimen­sional problems. Major areas of appli­ca­tion include econo­metrics, engi­nee­ring, biome­dical science, signal proces­sing and finance.
We consider the general problem of esti­ma­ting some specific quan­ti­ties of inte­rest of a generic copula (such as, for example, tail depen­dence index or the Spearman's coef­fi­cient) by adop­ting an appro­xi­mate Baye­sian approach based on compu­ting the empi­rical likelihood as an appro­xi­ma­tion of the likelihood func­tion for the quan­tity of inte­rest.
The approach is general, in the sense that it could be adapted both to para­metric and nonpa­ra­metric model­ling of the marginal distri­bu­tions and on a para­metric or semi­pa­ra­metric esti­ma­tion of the copula func­tion. We will show how the Baye­sian proce­dure based on ABC shows better proper­ties that the clas­sical infe­ren­tial solu­tion avail­able in the lite­ra­ture and apply the method in both simu­lated and real exam­ples.


Kindly note that on November 16 two talks are sche­duled at our insti­tute:
9:00 – 10:10  Clara Grazian (Univer­sity of Oxford)
11:00 – 12:10  Christa Cuchiero (Univer­sity of Vienna)



zurück zur Übersicht