WEBINAR: Longitudinal compositional data analysis (A. Panarotto; 20250515)

CoDa Association

WEBINAR

 

A couple of models for longitudinal compositional data analysis

Andrea Panarotto

To get access to the video recording of the talk, please, login and go to menu

Members area -> Video Recording & slides

---------------------------------------------------------------------------------------------------

 

  • Day: 15-05-2025
  • Time: 12:00h (CEST: UTC/GMT +2 h)
  • Link: https://us06web.zoom.us/j/82386133032?pwd=xkWeqwK10Vv6Xfu5LeH2ahFJifOb7d.1
  • Speaker: Andrea Panarotto (Department of Statistical Sciences, University of Padova, Italy)
  • Title: A couple of models for longitudinal compositional data analysis
  • Moderator: Gianna Monti
  • Abstract: In longitudinal compositional data analysis, tailored methods are required for capturing the dependency between successive compositional observations while taking into account the constrained nature of the data. In this talk, we introduce two novel methodologies, with different premises and goals, to address the task. In the first work, driven by a sustainable mobility motivation, we propose a state-space approach for the representation and the model-based clustering of compositional time series. To uncover the main patterns of urban mobility, the method is applied to real human movement data, where people’s trajectories are represented in the simplex by the proportions of road types in their surroundings. In the second work, we provide an alternative approach, by modeling the observations directly on the simplex and the longitudinal aspect through generalized estimating equations. This more general model is useful for applications where the dependence on the covariates is the main interest.
  • Keywords:  Aitchison geometry, longitudinal data,  time series  

 

---------------------------------------------------------------------------------------------------

Please note that the webinar will be recorded. Feel free to forward this message to your colleagues who may be interested in attending.

Check your local time at https://www.timeanddate.com/worldclock/converter.html?iso=20250515T100000&p1=31&p2=179&p3=152&p4=101