[Statlist] Research Seminar in Statistics | *FRIDAY 27 October 2023* | GSEM, University of Geneva

gsem-support-instituts g@em-@upport-|n@t|tut@ @end|ng |rom un|ge@ch
Mon Oct 23 09:52:18 CEST 2023


Dear all,



We are pleased to invite you to our next Research Seminar, organized by Professor Sebastian Engelke on behalf of the Research Center for Statistics

< https://www.unige.ch/gsem/en/research/institutes/rcs/team/ >.



FRIDAY 27 OCTOBER 2023, at 11:15 am, Uni Mail M 5220 & Online

Zoom research webinar: https://unige.zoom.us/j/92924332087?pwd=U1U1NFk4dTFCRHBMeWYrSDBQcXBiQT09

Meeting ID: 929 2433 2087

Passcode: 399192


Robust Diagonalwise Multivariate Singular Spectrum Analysis

Dr. Fabio CENTOFANTI, University of Naples Federico II, Italy

< https://www.researchgate.net/profile/Fabio-Centofanti >



ABSTRACT:
Multivariate Singular Spectrum Analysis (MSSA) serves as a widely adopted nonparametric technique for multivariate time series analysis. Nonetheless, MSSA faces challenges in handling outliers due to its reliance on singular value decomposition, which makes it sensitive to unusual data points. Consequently, MSSA can produce biased outcomes leading to potentially erroneous conclusions. In this seminar, a new variant of MSSA is introduced, named RObust Diagonalwise Estimation of SSA (RODESSA), designed to enhance its robustness against both cellwise and casewise outliers. Specifically, RODESSA replaces the decomposition step of MSSA with a robust low-rank approximation of the trajectory matrix, accounting for its unique structural characteristics. Additionally, a novel graphical representation is introduced, named the enhanced time series plot, to aid in visualizing different types of outliers. To assess its effectiveness, an extensive Monte Carlo simulation study is conducted, comparing RODESSA with competing methods from the existing literature. Furthermore, the practical applicability of our approach is illustrated through a real-world example involving the analysis of temperature data in passenger railway vehicles.



> View the Research Seminar agenda: < https://www.unige.ch/gsem/en/research/seminars/rcs/ >



Regards,



Laura Ferreira da Silva

Assistant to research and institutes

mailto:gsem-support-instituts using unige.ch


	[[alternative HTML version deleted]]



More information about the Statlist mailing list