This talk will tell the story of the many ideas and developments that emerged from a collaboration between marine ecologists and statisticians. Over the last 3 years, we have collaborated on the study of a benthic population of sea cucumbers (holothurians) near the coast of Giglio Island (Tuscany region). In the general framework of Bayesian point process modeling, we had to face:
1. complex sampling strategies (and these are presence-only data)
2. different strategies to build the dataset (automatic versus manual)
3. space-time variation
4. model validation
and more.
Point 2 allows for integration between classical statistical modeling and machine learning techniques. Point 4 is of particular interest, as developing cross-validation-type procedures for dependent data and point processes are still very active areas of development.
29 Maggio 2026, ore 12
Giovanna Jona Lasinio
Dept of Statistical Science, Sapienza University of Rome
Joint work with Gianluca Mastrantonio (Politecnico di Torino), Alessio Pollice (Università di Bari), Daniele Ventura (Sapienza, University, Rome), Gian Mario Sangiovanni (Sapienza University, Rome), Daniele Poggio (Politecnico di Torino)
In person: Room V (4th floor) building CU002 Scienze Statistiche
Webinar: https://uniroma1.zoom.us/j/83625004899?pwd=bXCtz0mp759PUh2lkqT0BUoVa0Uegg.1
ID riunione: 836 2500 4899
Passcode: 123456